WEBVTT
00:00:00.000 --> 00:00:11.420
Welcome to Episode 438 of the Microsoft Cloud IT Pro Podcast, recorded live on October 2, 2026.
00:00:11.960 --> 00:00:21.280
This is a show about Microsoft 365 and Azure from the perspective of IT pros and end users, where we discuss a topic, more recent news, and how it relates to you.
00:00:21.280 --> 00:00:32.520
Back in Episode 429, Scott and Ben introduced LLM wikis. Today, they're circling back around to see what they've turned into after using them daily for a few months.
00:00:32.860 --> 00:00:42.820
Scott walks through the skills that now run his daily life, from a morning prioritization pass and meeting prep to meeting capture and end-to-end daily reviews.
00:00:42.820 --> 00:00:50.660
We also talk about growing a workflow from an on-demand skill to a scheduled skill, loops, and deterministic scripts.
00:00:51.020 --> 00:00:59.860
We also discuss different approaches Scott and Ben have to creating custom dashboards and command centers and how they interact with their data.
00:01:01.880 --> 00:01:07.460
We're live with another episode of the Microsoft Cloud IT Pro Podcast, live from our houses.
00:01:07.740 --> 00:01:09.380
Live from New York. It's Saturday night.
00:01:09.640 --> 00:01:11.160
Oh, no, that's a different one.
00:01:11.280 --> 00:01:13.700
Different show. We're not quite that popular, Scott.
00:01:13.840 --> 00:01:14.160
Someday.
00:01:14.160 --> 00:01:20.480
I know we look that good, and we sound that good, and we could clearly do a primetime, not really primetime, late night,
00:01:20.760 --> 00:01:24.580
NBC show. Better be careful. We're going to end up talking about copyrighted stuff.
00:01:24.640 --> 00:01:28.260
We could go see if we could end up on The Onion and the new InfoWars channel or something.
00:01:28.440 --> 00:01:34.440
We could just spin yarn and conspiracy theories about Azure and Microsoft 365.
00:01:34.740 --> 00:01:38.100
365. We could. We could come up with all kinds of conspiracy theories.
00:01:38.200 --> 00:01:40.280
My daughter is actually enjoying that right now.
00:01:40.340 --> 00:01:40.980
Do you know it'll be fine?
00:01:40.980 --> 00:01:50.920
Yes, my daughter's actually taking a class right now that's all about conspiracy theories, where they spend like a week, because she's homeschooled.
00:01:50.920 --> 00:02:01.320
So all the kids in the class go home from their co-op, spend a week like looking at different conspiracy theories, and then come back to class the next week and debate the various conspiracy theories.
00:02:01.720 --> 00:02:09.960
So it's kind of interesting. She's having fun. They did the moon landing today. So the last week, she's been looking at all the different conspiracy theories around the moon landing.
00:02:09.960 --> 00:02:25.600
My oldest son is, he's studying history in college. And so he's, so he's in his second year now when he's getting kind of like more into things. And he has had the realization that history repeats itself.
00:02:27.000 --> 00:02:32.300
You didn't learn this in like middle school or high school, but you're coming to it in college. Okay, well, good realization.
00:02:32.680 --> 00:02:35.620
I mean, at least he came to that realization.
00:02:35.620 --> 00:02:40.120
There's some maturity there. I have hope that he'll move out of my house someday. We'll see.
00:02:40.880 --> 00:03:05.540
So you know what else happened today, Scott? I called my HVAC company because we have like in Florida, AC is kind of important. So twice a year, we have like regular, we have a plan. We sign up for discounted maintenance and all that. But then they do like twice annual, twice annual maintenance checks. And I got a text message and an email and all of it. Hey, it's time to schedule one.
00:03:05.540 --> 00:03:23.060
So I called them today because usually that works better than trying to do it online and all of that. And it was like, you're on hold. And then I had someone picked up the phone and it was like, I can't remember the name. Hi, I'm Lisa. This is your AI assistant. Like it was an AI.
00:03:23.060 --> 00:03:35.260
Well, he should announce itself that way. Person. I'd rather fall into like a phone tree and interactive voice recording like that, where it's very upfront and clear that yes.
00:03:35.260 --> 00:03:46.300
Yeah. And that's what it did. Name. But it was a little, it was interesting. I wanted to be annoyed at it the whole time because I knew I was talking to AI, but it actually went really smooth.
00:03:46.460 --> 00:04:00.360
They superimposed normal office sounds in it. So I could hear like a keyboard clicking in the background and all the stuff. But this AI assistant was, why are you calling? I was like annual maintenance and asked me all the questions it needed.
00:04:00.360 --> 00:04:14.920
Like had an actual conversation with this AI agent, just like I would have had with anybody in their office and was able to schedule my appointment, got the text message and the email with the confirmation of the appointment date and time.
00:04:14.920 --> 00:04:28.020
And as much as I wanted to be annoyed knowing I was talking to AI, it actually was a good experience. And I was pleasantly surprised and impressed with how well and smoothly it went.
00:04:28.020 --> 00:04:35.940
Good to hear. When you were texting me about it and telling me about it, I thought you were going to tell me like you fired off because there's lots of personal agents out there right now.
00:04:36.120 --> 00:04:45.540
So you've got Meadow with Muse, there's some invite stuff like Instinct that's out there. OpenAI announced Dots at their dev day this week.
00:04:45.960 --> 00:04:55.200
So I thought what you were going to tell me was like, oh, you had your AI personal assistant, because now these things are starting to come with phone numbers, and do inbound outbound calling.
00:04:55.200 --> 00:05:04.680
You had your AI assistant call another AI assistant, and maybe they got caught in one of those loops of like talking to each other in like binary or like whatever.
00:05:05.300 --> 00:05:06.440
No, but that would be funny.
00:05:06.780 --> 00:05:12.060
Sound and wavelengths. They wanted to have like a waveform representation to go and talk back and forth.
00:05:12.220 --> 00:05:16.500
I should try that next time is have my bot call their bot.
00:05:16.720 --> 00:05:25.100
Speaking of bots and Dots, did you also see the somewhat trolling, hilarious news like OpenAI announced their Dots?
00:05:25.100 --> 00:05:31.140
And immediately after that, Musk went out and bought Dot.com and redirected it to the GrokBot.
00:05:34.480 --> 00:05:38.740
If you want a domain name about a product, you're going to buy the domain name before you announce it.
00:05:38.820 --> 00:05:43.080
Or you just get big enough that you have your own company domain name, right?
00:05:43.160 --> 00:05:46.160
Like the whole like Cloud, Dot, Microsoft, and things like that.
00:05:46.720 --> 00:05:47.860
See, that would solve it too.
00:05:47.960 --> 00:05:49.800
That's the other way to go about it.
00:05:49.920 --> 00:05:52.600
But yeah, we should talk about AI stuff today.
00:05:52.760 --> 00:05:53.240
We should.
00:05:53.240 --> 00:06:09.540
I want to revisit the talk that we had a little while ago about getting started with LL and Wikis and kind of talking about what they can turn into with a little bit of maturity.
00:06:10.080 --> 00:06:14.920
So like this will be a little bit more of like a freeform conversation for us.
00:06:15.100 --> 00:06:16.760
I want to talk about my setup.
00:06:17.060 --> 00:06:20.040
I want to learn a little bit more about your setup.
00:06:20.640 --> 00:06:28.820
And I really want to dig into kind of the workflows and maybe like the skills and agents that you're using.
00:06:28.820 --> 00:06:39.600
Like what are the harnesses that you're using around this that make it helpful for you and empower your day-to-day work, accelerate your day-to-day work.
00:06:39.940 --> 00:06:41.160
Maybe they make things harder, right?
00:06:41.160 --> 00:06:42.840
As you've been experimenting and going through.
00:06:43.100 --> 00:06:46.980
So what are things that maybe you hit a brick wall with or I hit a brick wall with?
00:06:47.320 --> 00:06:52.060
And we kind of went in a little bit of a different direction.
00:06:52.620 --> 00:06:54.580
So yeah, you want me to go first?
00:06:54.680 --> 00:06:55.040
Sure.
00:06:55.160 --> 00:06:55.640
What I've been doing?
00:06:55.780 --> 00:06:56.060
Yes.
00:06:56.200 --> 00:06:57.640
I'm curious to hear what you've been doing.
00:06:58.100 --> 00:07:02.420
So I don't want to get to your point earlier about like copyright infringement, things like that.
00:07:02.580 --> 00:07:03.620
I don't want anything there.
00:07:03.620 --> 00:07:17.220
But I took what were like various instantiations of LL and Wikis that I've been working with around like various areas of like research for specific projects and initiatives that I work on.
00:07:17.320 --> 00:07:20.280
Just like my day-to-day tasks, tracking, things like that.
00:07:20.660 --> 00:07:27.520
And I consolidated them all into one wiki and it's my WorkOS, right?
00:07:27.900 --> 00:07:29.420
It's my work operating system.
00:07:30.040 --> 00:07:37.140
So it's meant to track really everything that I do across work, whether that's working.
00:07:37.560 --> 00:07:41.460
So I'm a product manager, so that could be working on a new product.
00:07:41.740 --> 00:07:43.940
It could be working on an update to an existing feature.
00:07:44.180 --> 00:07:46.160
It could be like building a new feature.
00:07:46.320 --> 00:07:47.980
Maybe that's not a full product, right?
00:07:47.980 --> 00:07:51.560
But it's complimentary to existing product suite, things like that.
00:07:51.560 --> 00:08:08.140
And I do all the PME things that come along with that of market research, writing PRDs, specs, having customer conversations, wrangling work management systems like ADO.
00:08:08.560 --> 00:08:09.640
We use Azure DevOps.
00:08:10.380 --> 00:08:13.040
Maybe somebody's using JIRA, something like that.
00:08:13.040 --> 00:08:23.580
But really, everything that I do gets logged, gets tagged, and goes into that WorkOS.
00:08:24.300 --> 00:08:28.700
So off the cuff, that sounds like, okay, but how do I get everything in there?
00:08:29.280 --> 00:08:33.140
And what's important and what happens along the way?
00:08:33.780 --> 00:08:44.080
So I've got a set of skills that are out there that I've put together that work for my way of work and that operate within the harness that I use.
00:08:44.340 --> 00:08:50.380
So the harness that I use today is Microsoft Scout, the desktop app for Scout.
00:08:50.740 --> 00:08:55.380
I'm going to be a little bit sad when that goes to autopilot here in the future because I don't know what that's going to do.
00:08:55.560 --> 00:08:57.820
Not autopilot because we have Microsoft autopilot.
00:08:58.000 --> 00:09:01.400
We need to differentiate now, Scott, because we have Microsoft autopilot.
00:09:01.560 --> 00:09:02.880
Autopilot inside of copilot.
00:09:03.000 --> 00:09:03.800
Copilot autopilot.
00:09:03.900 --> 00:09:05.180
And copilot autopilot.
00:09:05.520 --> 00:09:07.540
That's another topic for another day.
00:09:07.660 --> 00:09:08.680
We'll get to that one eventually.
00:09:09.400 --> 00:09:23.780
So I've got this set of skills that help me with inputs and outputs to get information into that wiki and then help me get good concrete outputs out of it.
00:09:23.780 --> 00:09:33.760
So one would be, for example, one of the ones I use quite a bit actually is an email thread ingester skill.
00:09:34.380 --> 00:09:41.680
So my work takes place across teams, email, in-person conversations, teams meetings.
00:09:41.680 --> 00:09:44.700
If I'm meeting with customers, it could be Zoom meetings, things like that.
00:09:45.200 --> 00:09:53.000
But particularly for internal conversations, a lot of that stuff still happens in email or there's newsletters and various things that go out.
00:09:53.660 --> 00:09:57.960
So I have a little skill that's my WorkOS email thread ingester.
00:09:59.040 --> 00:10:09.320
And because it works in context of WorkIQ and the Microsoft Graph, it respects sensitivity labels, DLP, all those kinds of things.
00:10:09.320 --> 00:10:17.920
It's not like I'm taking super secret stuff out of the system and throwing it over in this random text wiki over here that lives in my OneDrive or anything like that.
00:10:18.300 --> 00:10:19.520
So I use that one quite a bit.
00:10:19.880 --> 00:10:24.420
And that's super handy and helpful for me for point-in-time kinds of things.
00:10:25.160 --> 00:10:32.960
I have a skill that runs every single morning for me that goes every single weekday for me.
00:10:32.960 --> 00:10:41.420
It runs every morning at 9 a.m. local, which is my WorkOS daily prioritization skill.
00:10:42.220 --> 00:10:48.000
So what that does is it looks at the to-dos that are in my WorkOS.
00:10:48.340 --> 00:10:55.080
So I made this very conscientious decision not to use Planner or Microsoft to-do or anything like that.
00:10:55.220 --> 00:10:59.320
I'm just kind of managing everything in markdown files inside my WorkOS.
00:10:59.620 --> 00:11:04.360
So your source of truth for your tasks is these markdown files in WorkOS?
00:11:04.640 --> 00:11:04.900
Yes.
00:11:05.100 --> 00:11:05.320
Okay.
00:11:05.420 --> 00:11:08.080
But I find that works because I can tag them.
00:11:08.500 --> 00:11:10.220
They carry the tags through the wiki.
00:11:10.620 --> 00:11:13.720
They carry all the grounding and the context.
00:11:14.000 --> 00:11:22.400
I can link to-dos very easily to projects or to initiatives and various other contexts that exist in that wiki.
00:11:22.640 --> 00:11:38.040
So what my kind of daily prioritization skill does for me is it's meant to look at my prior day and or my incomplete tasks that are sitting out there that I've already spec'd out and help me kind of rationalize,
00:11:38.440 --> 00:11:51.800
here's the number one thing that you should do today out of your existing to-do list based on what happened between yesterday and today through your email, through your team signals, all those kinds of things, documents people shared with you.
00:11:52.220 --> 00:11:55.080
Here's another task that you should go and do.
00:11:55.580 --> 00:12:04.900
And then if it has room for it, the way I structured the skill is I always want to have kind of like three things on my plate in front of me that are like these are like the must-dos for today kind of thing.
00:12:05.420 --> 00:12:06.280
So that sits out there.
00:12:06.280 --> 00:12:09.400
I have a question about that one because I have one similar to that.
00:12:09.740 --> 00:12:30.020
Have you, and this is something I've done with mine with some success, have you had that skill when it looks at your daily to-dos, also look at your meetings for the upcoming day to prioritize if there's anything that you have to have done for a meeting coming up in like the next eight hours based on your tasks and projects as well?
00:12:30.200 --> 00:12:30.780
Yes and no.
00:12:30.940 --> 00:12:33.440
So we should get into that one next.
00:12:33.520 --> 00:12:33.740
All right.
00:12:33.740 --> 00:12:40.580
I want to kind of start the conversation here talking about the skills that I've integrated and that I'm using within Scout as a harness.
00:12:40.860 --> 00:12:45.040
I could be using Copilot, CLI, Claude, Codex, Anti-Gravity.
00:12:45.140 --> 00:12:45.900
It really doesn't matter, right?
00:12:46.000 --> 00:12:51.680
A skill is a skill and that markdown is portable and can come over and I can tie my connectors, all that kind of stuff.
00:12:52.200 --> 00:13:02.320
But there's this concept that I think becomes very powerful of moving from skills to loops and having loops that go out and execute things for me.
00:13:02.680 --> 00:13:07.320
So while like my daily prioritization skill is just a skill, I could run it on demand.
00:13:07.560 --> 00:13:09.660
I could come in and I could say, hey, run this thing.
00:13:10.080 --> 00:13:12.820
It's both a skill and it's a loop for me.
00:13:13.180 --> 00:13:18.680
So I put it in the harness and I schedule it and I say, hey, run this every day at 9 a.m.
00:13:18.680 --> 00:13:30.260
I actually have, I can run the same skill like every 30 minutes basically on a constant loop and have it do constant reevaluation for me like throughout the day and pivot things through.
00:13:30.260 --> 00:13:44.620
So I have two other skills when it comes to meetings that also function as skills and or more harnessed loops and kind of just on the flywheel of they're going to do things automatically for me.
00:13:44.980 --> 00:13:48.160
So one is a meeting prep skill.
00:13:48.540 --> 00:13:54.340
So I have like WorkOS-MeetingPrep and I have WorkOS-MeetingCapture.
00:13:54.600 --> 00:13:57.820
So WorkOS-MeetingPrep is my prep skill.
00:13:57.820 --> 00:14:16.640
So as my daily prioritization skill is running in the morning and it's going through like that harness and that automated scheduling and now it's turning more into a loop, what that's actually doing, one of the subtasks that it has is it goes and runs meeting prep for every single meeting that I have that day.
00:14:16.700 --> 00:14:18.100
That's on my calendar, okay?
00:14:18.680 --> 00:14:27.040
And then as I'm coming out of meetings, many of the meetings we attend are recorded, they have transcripts, they have all of those things.
00:14:27.040 --> 00:14:28.700
I have a meeting capture skill.
00:14:29.240 --> 00:14:38.520
And my meeting capture skill goes and evaluates the recording, it evaluates the transcript, it grabs the WorkIQ context, which is inclusive of like the meeting chat.
00:14:38.760 --> 00:14:42.160
So you can catch like documents that were shared in the meeting, things like that.
00:14:42.660 --> 00:14:50.500
And that comes through and effectively it turns my meetings into a two-step flow of prep and then capture.
00:14:50.860 --> 00:14:58.340
Or if I have a meeting that maybe, let's say like I was an optional attendee or maybe it was a learning forum, it's something that I want to catch up on later.
00:14:58.780 --> 00:15:00.000
So I wasn't active in it.
00:15:00.300 --> 00:15:09.020
For like meetings on my calendar that I'm just following or I'm an optional attendee, things like that, my morning prioritization skill doesn't actually run prep for those automatically.
00:15:09.140 --> 00:15:10.060
It just skips over them, right?
00:15:10.060 --> 00:15:11.340
I don't need them right now.
00:15:11.740 --> 00:15:21.900
So for those, if I want to, I can go in and I can basically run, I kind of have this like overarching maybe superset skill, which is generate and capture.
00:15:22.280 --> 00:15:29.160
So I can run generate individually, I can run capture post-meeting, or I can run generate and capture, which kind of does both for me.
00:15:29.360 --> 00:15:36.840
And generating capture is helpful for the meeting that's already happened that I didn't attend because they do different things, right?
00:15:36.940 --> 00:15:43.860
Prep goes out and looks at my inbox, it looks at conversations with meeting attendees, and it says, hey, what's the context for this thing?
00:15:44.280 --> 00:15:54.580
Capture is go blend that kind of meeting brief and the meeting prep with the actual outcomes of that meeting and what's next along the way.
00:15:55.240 --> 00:16:04.340
And then I have another skill that runs in a loop, again, on an automation, which is my end-of-day wrap-up.
00:16:04.640 --> 00:16:12.260
And part of my end-of-day wrap-up is that skill evaluates my wiki and it says, what are the meetings I attended today that I forgot to capture?
00:16:12.260 --> 00:16:19.560
Or I forgot to do a prep and capture on, and it just goes and grabs them automatically for me and kind of brings that context in.
00:16:19.960 --> 00:16:27.240
And that's good for me because then in that virtuous cycle of tomorrow's a new day, my automation for my daily prep runs again.
00:16:27.700 --> 00:16:31.080
And now it's able to catch up on the meetings and the context and things that are there.
00:16:31.420 --> 00:16:40.160
So I don't have to constantly be like human intervention in the loop to say, like, click a button or go run this skill, go run this skill, go run this skill in my harness.
00:16:40.440 --> 00:16:48.560
It's just happening kind of automatically and automagically as I go through my day.
00:16:49.240 --> 00:17:00.160
And then, let's see, because this is an LLM wiki, like, if people are on, like, the Carpathi-style LLM wiki path, they often end up with, like, that raw or clippings or ingest folder for things.
00:17:00.540 --> 00:17:01.620
Well, I'm a PM.
00:17:01.760 --> 00:17:02.400
I do research.
00:17:02.520 --> 00:17:04.380
I have to keep up on industry news.
00:17:04.740 --> 00:17:08.500
It could be Gartner, IDC, launch from a competitor, things like that.
00:17:08.760 --> 00:17:12.580
So I'm just running things in a markdown wiki at Subsidion.
00:17:12.800 --> 00:17:14.560
So I can use Subsidion Web Clipper.
00:17:14.980 --> 00:17:18.340
So I send things into a, like, a raw or a clippings directory.
00:17:18.880 --> 00:17:21.600
So I have a, hey, go ingest my clippings.
00:17:21.880 --> 00:17:23.460
Like, go rationalize over this.
00:17:23.820 --> 00:17:26.120
And again, organize it in context of my projects.
00:17:26.120 --> 00:17:26.740
Tag it.
00:17:27.500 --> 00:17:31.060
Do all those things that you need to do along the way.
00:17:31.700 --> 00:17:34.120
And then I have these kind of, I don't run them in loops.
00:17:34.300 --> 00:17:35.060
I don't run them automatically.
00:17:35.280 --> 00:17:39.800
I have a bunch of, I call them, like, my little, like, sub-agents or my helper skills.
00:17:40.700 --> 00:17:44.760
So I'm responsible for a portion of a business.
00:17:44.940 --> 00:17:48.060
That business has, like, a monthly business review.
00:17:48.580 --> 00:17:50.300
We do ongoing monthly product reviews.
00:17:50.600 --> 00:17:53.520
We do weekly execution reviews, things like that.
00:17:54.040 --> 00:18:00.480
And those often require, like, very structured inputs of, let's go grab your ADO items, get the current status.
00:18:00.480 --> 00:18:03.520
What's the dates, what's the update, things like that.
00:18:04.100 --> 00:18:12.920
So I have these skills that go out and help me do those things, like WorkOS MBR prep, WorkOS product review prep, things like that.
00:18:13.480 --> 00:18:21.120
And they all go out and they kind of work and they help me get the rough outlines for those things.
00:18:21.380 --> 00:18:23.480
I don't let them, like, write everything for me.
00:18:23.820 --> 00:18:28.100
I have this very visceral reaction to, like, just pure, like, AI-written things.
00:18:28.280 --> 00:18:37.840
I'll let them bullet it out so that then I can take them and kind of put my own thinking or my own spin on top of it along the way.
00:18:38.160 --> 00:18:47.100
And then the last thing that I did, and this is probably, like, super niche, but it's been very effective for me,
00:18:47.340 --> 00:18:52.560
is I took the things that I was doing in the harness.
00:18:53.380 --> 00:18:55.620
So I started in Copilot CLI, eventually moved to Scout.
00:18:55.820 --> 00:18:57.280
Now Scout's kind of like sitting over there.
00:18:57.620 --> 00:18:59.620
So Scout's really like my automation engine.
00:18:59.940 --> 00:19:02.280
It's not actually where I go and run these things from.
00:19:02.640 --> 00:19:09.020
So if I need to go, say, run a meeting prep, so I've got an upcoming meeting and that prep hasn't been run yet,
00:19:09.100 --> 00:19:13.520
or I've got to go run a capture, I built myself a WorkOS dashboard.
00:19:14.020 --> 00:19:24.380
So I have a little local HTML site that's served to me, and that site presents to me my calendar.
00:19:24.960 --> 00:19:29.360
It presents the meetings that haven't been prepped versus those that have been prepped,
00:19:29.480 --> 00:19:31.940
those that haven't been captured versus those that haven't been captured.
00:19:32.380 --> 00:19:34.280
It gets my to-dos there for me.
00:19:34.280 --> 00:19:40.720
So rather than me having to go into Obsidian, say I'm updating a to-do from not started to in progress,
00:19:41.180 --> 00:19:44.420
I actually, I go to this website, and I just click the to-do.
00:19:44.580 --> 00:19:47.060
It's kind of like to-doist or something for me just sitting there.
00:19:47.200 --> 00:19:51.680
And I update it right there, and that thing goes and writes the wiki, the markdown for me.
00:19:51.760 --> 00:19:54.340
So I don't have to go, like, find the markdown file and worry about, like,
00:19:54.340 --> 00:19:55.800
oh, did I get the tag right, everything.
00:19:56.300 --> 00:19:58.980
Things are very deterministic for me.
00:19:59.600 --> 00:20:03.420
So I've got the wiki there when I need it, when I need to go search for it.
00:20:03.480 --> 00:20:04.560
I do all that through Obsidian.
00:20:04.840 --> 00:20:09.040
I've got Scout sitting on the side, which is running my loops and running my automations
00:20:09.040 --> 00:20:13.740
and is kind of the house for those skills in the execution engine.
00:20:14.140 --> 00:20:17.740
And then I've got this just kind of dashboard that I live in.
00:20:18.060 --> 00:20:20.880
And from the dashboard, I can do things, like, I've got, like, my calendar in there.
00:20:20.880 --> 00:20:24.760
Like, if I went and looked at my agenda view today, it would have this podcast recording
00:20:24.760 --> 00:20:25.520
as a meeting on there.
00:20:25.920 --> 00:20:28.880
I can click on podcast recording, and then I've got a little button there that says,
00:20:29.360 --> 00:20:32.280
open your brief, and it'll just pop me open straight into Obsidian.
00:20:32.680 --> 00:20:39.400
Like, it takes me right into the vault, right to that thing, and gets me ready to go and
00:20:39.400 --> 00:20:40.700
where I need to be.
00:20:41.300 --> 00:20:44.800
So that's been super helpful and super contextual for me.
00:20:44.820 --> 00:20:49.540
Because when I come in the morning, I refresh my homepage for my WorkOS, right?
00:20:49.540 --> 00:20:51.520
I just hit the refresh button.
00:20:52.040 --> 00:20:54.860
And then all of a sudden, I'm presented with, here's your daily priorities.
00:20:55.260 --> 00:20:56.940
Here's your top three to-dos for the day.
00:20:57.160 --> 00:20:58.620
Here's your agenda for the day.
00:20:59.020 --> 00:21:02.540
Here's how these priorities relate to your agenda for the day.
00:21:02.760 --> 00:21:04.720
Here's all the things that I generated for you.
00:21:04.820 --> 00:21:06.040
Here's how many aren't left.
00:21:06.400 --> 00:21:06.860
Things like that.
00:21:07.260 --> 00:21:09.880
It's a very freeing thing.
00:21:10.160 --> 00:21:13.020
Like, in the past, I would have sat there and said, like, oh, crap, I've got a meeting.
00:21:13.180 --> 00:21:13.980
Let me go to that meeting.
00:21:14.400 --> 00:21:18.080
And you kind of come in all discombobulated, and you do the context switch, and you come
00:21:18.080 --> 00:21:18.660
out to the other side.
00:21:18.960 --> 00:21:20.600
Then you go, okay, I've got to go to my next meeting.
00:21:20.700 --> 00:21:24.500
I don't have time to go back and kind of write my notes up or capture them, anything
00:21:24.500 --> 00:21:25.060
like that.
00:21:25.440 --> 00:21:27.440
Now it's all just turnkey.
00:21:27.920 --> 00:21:32.820
It's all right there for me and ready to go.
00:21:33.280 --> 00:21:35.640
And also, very deterministic.
00:21:36.280 --> 00:21:42.060
So the last time that we chatted about this, I believe I had also mentioned kind of taking
00:21:42.060 --> 00:21:48.960
skills that maybe start out with a bunch of overarching shape in them, but then you want to make them
00:21:48.960 --> 00:21:50.540
more deterministic over time.
00:21:50.540 --> 00:21:54.680
So sometimes you have to stop relying on the model and the skill itself.
00:21:55.260 --> 00:22:00.220
And you need to kind of write scripts and have a script do the generation for you so that
00:22:00.220 --> 00:22:03.420
things are always formatted consistently, they come in the right way, all that.
00:22:04.080 --> 00:22:08.100
So for me, my meeting prep skill, it started as a very generic skill.
00:22:08.360 --> 00:22:13.380
Like, go find the transcript and the recording for this meeting on my calendar and capture it.
00:22:13.640 --> 00:22:16.480
And then I refined it over time, but it was still a bunch of markdown.
00:22:16.480 --> 00:22:21.760
I actually took it and worked and iterated on it back and forth a little bit, and I was able
00:22:21.760 --> 00:22:22.180
to get it.
00:22:22.360 --> 00:22:27.620
So now my meeting prep skill is really a script that calls WorkIQ.
00:22:28.120 --> 00:22:33.960
And it calls WorkIQ in a way that the output is very deterministic for me for every single
00:22:33.960 --> 00:22:34.740
meeting I go to.
00:22:34.940 --> 00:22:39.840
It's got the meeting attendees, it's got the transcript overview and the summary, it's got
00:22:39.840 --> 00:22:43.960
here's the key items from the chat, here's the documents that were shared, things like that.
00:22:44.520 --> 00:22:50.580
And what that allowed me to do by putting that skill, and now the skill is just a wrapper for
00:22:50.580 --> 00:22:53.800
calling that script, back to that dashboard thing, right?
00:22:53.840 --> 00:22:57.340
That dashboard's there, like I can't click a button on that dashboard and have it go run
00:22:57.340 --> 00:22:57.980
a skill.
00:22:58.240 --> 00:23:00.080
But I can have it go run a script, right?
00:23:00.120 --> 00:23:04.060
And kick that off very easily, like it just goes and runs a Python script in the background.
00:23:04.740 --> 00:23:09.880
So whether I'm running the skill through the harness, through Scout or through GitHub Copilot
00:23:09.880 --> 00:23:14.680
CLI, or I'm clicking the button, at the end of the day, the skill and me clicking the button
00:23:14.680 --> 00:23:20.480
are calling the same Python script, giving me the same deterministic output every single
00:23:20.480 --> 00:23:20.800
time.
00:23:20.880 --> 00:23:24.780
It's interesting to hear you talk because I've done some things very similar, but then
00:23:24.780 --> 00:23:26.220
I've also done some different things.
00:23:26.340 --> 00:23:28.620
Like I said, it's a very empowering thing, I think.
00:23:28.760 --> 00:23:32.460
So this is in no way, shape, or form a badge of honor.
00:23:32.800 --> 00:23:35.860
I'm actually kind of like scared when I say stuff like this out loud.
00:23:36.220 --> 00:23:39.920
I have on average 65 meetings a week on my calendar.
00:23:40.040 --> 00:23:40.800
That's nuts.
00:23:40.940 --> 00:23:41.860
So that's my average.
00:23:42.340 --> 00:23:49.020
Some weeks I have 80, some weeks I have 40, but I'm sitting right around 65 on average
00:23:49.020 --> 00:23:51.720
over the past three months since I started tracking this.
00:23:52.220 --> 00:23:55.380
I do not attend 65 meetings a week, to be very clear.
00:23:55.520 --> 00:23:59.100
Like 20 of those are probably just like learning or like FYI things.
00:23:59.600 --> 00:24:04.100
But having that skill out there, that prep and capture skill, like that still allows me to
00:24:04.100 --> 00:24:05.680
ingest the knowledge and have it out there.
00:24:05.760 --> 00:24:12.100
Like we had a training last week on like GitHub Copilot CLI and like sub-agent usage and things
00:24:12.100 --> 00:24:12.360
in it.
00:24:12.440 --> 00:24:13.320
I was like, oh, that's cool.
00:24:13.420 --> 00:24:14.580
Like I really want to go catch up with that.
00:24:15.000 --> 00:24:17.700
In the past, I would have completely forgotten about that, right?
00:24:17.720 --> 00:24:19.340
I would have had it written down on a piece of paper.
00:24:19.340 --> 00:24:23.020
I would have had like a to-do and to-do to go watch the recording for that.
00:24:23.360 --> 00:24:25.260
And I never would have gotten to it or come back around.
00:24:25.720 --> 00:24:30.080
But because I have this like prep and capture thing, it actually can sit there and contextualize
00:24:30.080 --> 00:24:34.540
for me and not only get me the information I need, so I can go get the summary and decide
00:24:34.540 --> 00:24:36.480
like, oh, was that actually worth my time or not?
00:24:36.960 --> 00:24:43.560
But then it's got all the canonical links back to the full recap and everything that I need
00:24:43.560 --> 00:24:45.360
out there and that's ready to go.
00:24:45.360 --> 00:24:50.220
So I'm probably like on some further end of the pendulum swinging.
00:24:50.440 --> 00:24:54.080
Like I haven't quite found equilibrium there where I'm probably still capturing too much
00:24:54.080 --> 00:24:56.480
stuff and I'm bringing in too much context.
00:24:56.680 --> 00:25:00.400
But I'm still at the point even like a couple months into this where I'd rather have more
00:25:00.400 --> 00:25:01.040
than less.
00:25:01.680 --> 00:25:04.960
And I hope I can get to like the less is more kind of thing along the way.
00:25:05.400 --> 00:25:10.360
But when it's so easy to do, because the loops are there and the harness is there, I'm kind
00:25:10.360 --> 00:25:11.600
of inclined just to keep it running.
00:25:12.280 --> 00:25:16.940
And then, like I said, it makes a bunch of this stuff like more powerful for me, right?
00:25:16.980 --> 00:25:23.280
Because if I'm going and like we're working on maybe like a new like CLI component or we're
00:25:23.280 --> 00:25:29.600
doing an MCP thing or some kind of like agentic thing and now I've got all these training on
00:25:29.600 --> 00:25:32.820
how agents work and how they do stuff, like that all applies, right?
00:25:32.880 --> 00:25:36.160
That all comes back into what I need to do.
00:25:36.360 --> 00:25:43.140
So contextually, that stuff could be part of my research for a SPAC or a PRD like that
00:25:43.140 --> 00:25:44.580
something later.
00:25:45.160 --> 00:25:49.000
And then I have these aspirations to go build some more loops.
00:25:49.500 --> 00:25:50.620
And I very much want to do this.
00:25:50.860 --> 00:25:58.740
So for me, for example, one of the things that I like to do is I like to go on Reddit and
00:25:58.740 --> 00:26:02.120
LinkedIn to go find feedback from users.
00:26:02.720 --> 00:26:04.000
So I've got a couple different places.
00:26:04.540 --> 00:26:07.080
Azure Storage really doesn't show up on Hacker News for some reason.
00:26:07.400 --> 00:26:08.880
They must not think we're a startup or something.
00:26:08.980 --> 00:26:09.220
I don't know.
00:26:09.620 --> 00:26:15.680
But we get like decent feedback on forums like Microsoft Q&A and Reddit and things like that.
00:26:15.900 --> 00:26:20.480
So I would love to go and build like a user research loop that can help me go like research
00:26:20.480 --> 00:26:24.520
my product on like a daily or weekly or monthly basis, things like that might be a little
00:26:24.520 --> 00:26:25.200
bit harder for me.
00:26:25.320 --> 00:26:26.360
I saw just today.
00:26:26.480 --> 00:26:27.740
So today is October 2nd.
00:26:28.240 --> 00:26:33.300
I think I saw earlier that Reddit is going to shut down all their external search access
00:26:33.300 --> 00:26:34.940
and all their RSS feed access.
00:26:35.460 --> 00:26:38.640
So that's going to make it a lot harder for agents to get at the data that's in Reddit.
00:26:38.920 --> 00:26:40.140
So maybe it's a good thing.
00:26:40.240 --> 00:26:44.360
I haven't built that loop and skill yet to go and do that.
00:26:44.880 --> 00:26:50.200
But the whole thing of starting with skills, automating those skills, and then figuring out
00:26:50.200 --> 00:26:54.940
how to compound them into these loops that just kind of keep running until you get to the
00:26:54.940 --> 00:26:57.080
outcome you want, super powerful.
00:26:57.080 --> 00:27:03.120
If you're managing file storage today, you've probably felt the pressure.
00:27:03.460 --> 00:27:05.300
Refresh cycles are harder to plan.
00:27:05.520 --> 00:27:10.420
Costs and complexity are increasing, especially as file workloads scale across regions.
00:27:10.900 --> 00:27:13.540
That's why Microsoft customers trust Nasuni.
00:27:13.820 --> 00:27:20.000
As an Azure certified software, Nasuni provides a global file system designed for enterprise scale
00:27:20.000 --> 00:27:25.700
environments, keeping file data secure, governed, and accessible across Azure environments without
00:27:25.700 --> 00:27:27.440
added infrastructure complexity.
00:27:27.440 --> 00:27:33.680
It's the foundation to keep your unstructured data stable, even as your infrastructure evolves.
00:27:33.940 --> 00:27:39.880
To see how much you could save while keeping your initiatives on track, visit nasuni.com
00:27:39.880 --> 00:27:50.040
I have a few that I run regularly.
00:27:50.040 --> 00:27:54.080
I have not gotten too far into the loops yet.
00:27:54.340 --> 00:27:57.800
I have a lot of stuff that I still just run manually.
00:27:57.800 --> 00:28:03.680
I have a few things that I run as routines or scheduled jobs.
00:28:04.100 --> 00:28:08.920
But it's been, yeah, it's a lot of the way I would look at these LLM wikis.
00:28:09.040 --> 00:28:14.000
And even again, it's probably taking it further than what the LLM wiki did, like with your dashboard
00:28:14.000 --> 00:28:14.600
and HTML.
00:28:15.020 --> 00:28:20.380
I actually built my own custom Obsidian plugin that I have as a dashboard.
00:28:20.820 --> 00:28:23.620
So my dashboard sits right in Obsidian rather than a web page.
00:28:23.820 --> 00:28:24.520
How did you do that?
00:28:24.760 --> 00:28:25.640
I'm intrigued.
00:28:25.920 --> 00:28:30.880
I told Claude, so I actually told Claude to go build me a custom Obsidian plugin.
00:28:31.300 --> 00:28:32.720
And Claude has built the...
00:28:32.720 --> 00:28:34.040
Okay, built it as a plugin.
00:28:34.220 --> 00:28:35.420
So it renders in the canvas.
00:28:35.600 --> 00:28:37.100
It built it as a plugin.
00:28:37.400 --> 00:28:41.800
So it renders as a canvas, like down the left side of Obsidian where you have your different
00:28:41.800 --> 00:28:42.220
icons.
00:28:42.700 --> 00:28:44.900
I now have an icon for my dashboard.
00:28:45.180 --> 00:28:47.300
And my dashboard has a bunch of cards on it.
00:28:47.560 --> 00:28:53.140
So this is where similar type of concept with mine, only I combined a bunch of stuff.
00:28:53.140 --> 00:28:57.700
I have multiple LLM wikis in the same vault because I have one for work.
00:28:57.940 --> 00:28:59.480
I have one for some personal stuff.
00:28:59.780 --> 00:29:01.940
I have a separate podcast one in my vault.
00:29:02.220 --> 00:29:10.280
Because sometimes for me, the overlap between a topic we talk about on a podcast and a blog
00:29:10.280 --> 00:29:15.200
post, maybe I want to write for my website and a client that I'm working with, there's so
00:29:15.200 --> 00:29:16.300
much overlap there.
00:29:16.300 --> 00:29:22.100
I want them to be separate to an extent, but I also want to be able to correlate certain
00:29:22.100 --> 00:29:22.860
things across them.
00:29:23.100 --> 00:29:26.200
But then my dashboard has cards that are color-coded and labeled.
00:29:26.640 --> 00:29:34.840
So I have intelligent cards for my top to-dos, maybe things that are on fire, CRM type information,
00:29:35.140 --> 00:29:38.160
potential jobs, where jobs are in the pipeline.
00:29:38.480 --> 00:29:45.720
I have some podcast stuff where I pull in podcast stats so I can see trends with the podcast
00:29:45.720 --> 00:29:49.100
downloads and episodes month to month.
00:29:49.420 --> 00:29:52.380
I have some personal stuff where I actually pull in stock market.
00:29:52.380 --> 00:29:58.620
I have a markdown file where I put stock tickers I want to track, and then I have a dashboard
00:29:58.620 --> 00:30:03.520
that every time I go to it, it'll go out and pull the latest stock prices and put it on
00:30:03.520 --> 00:30:04.060
the dashboard.
00:30:04.540 --> 00:30:10.380
But then I put little buttons on it that'll call a skill to go do analysis on that stock
00:30:10.380 --> 00:30:10.700
ticker.
00:30:10.800 --> 00:30:16.020
So I have my own analysis on stock tickers within my LLM wiki.
00:30:16.160 --> 00:30:18.760
And I might have to explore this because one of the things...
00:30:18.760 --> 00:30:19.560
It's interesting.
00:30:19.740 --> 00:30:23.700
I am definitely juggling between Scout, which sits out there.
00:30:23.780 --> 00:30:24.700
It's a desktop app, right?
00:30:24.720 --> 00:30:25.480
It has to be running.
00:30:25.980 --> 00:30:28.960
So I've got Scout sitting out there running and...
00:30:28.960 --> 00:30:29.820
Wow, that hurt.
00:30:30.100 --> 00:30:30.780
Sorry about that.
00:30:30.920 --> 00:30:33.660
It was quite a whine by that dog, by that puppy.
00:30:33.780 --> 00:30:35.180
I've got Scout out there running.
00:30:35.660 --> 00:30:37.760
I have this website that kind of runs.
00:30:37.840 --> 00:30:39.320
It's just a pin tab in Edge for me.
00:30:39.320 --> 00:30:41.580
It's always my first tab sitting at the top.
00:30:42.200 --> 00:30:44.940
And then I have Obsidian.
00:30:45.440 --> 00:30:49.440
So having just one less thing, like I don't have to go to the website and I can just live
00:30:49.440 --> 00:30:49.880
in Obsidian.
00:30:50.240 --> 00:30:53.160
Because I do this thing where I've been tempted to...
00:30:53.160 --> 00:30:54.320
Maybe I was thinking about it wrong.
00:30:54.380 --> 00:30:59.400
I was actually tempted to see if I could find a way to embed a Markdown viewer in an Obsidian
00:30:59.400 --> 00:31:00.440
view in my website.
00:31:00.700 --> 00:31:02.200
But I should just totally go the other way.
00:31:02.720 --> 00:31:04.120
The Canvas thing is very interesting.
00:31:04.120 --> 00:31:06.060
So I'm glad you brought that up.
00:31:06.140 --> 00:31:09.360
I've seen a bunch of people doing similar things recently.
00:31:10.100 --> 00:31:16.340
So if you're in VS Code and you're using, say, like GitHub Copilot chat in VS Code.
00:31:16.960 --> 00:31:19.740
So VS Code has plugins, all those kinds of things.
00:31:19.840 --> 00:31:24.580
You can absolutely use your LLM to help you develop a plugin that draws on the canvas in
00:31:24.580 --> 00:31:25.160
VS Code.
00:31:25.520 --> 00:31:27.540
The other one is GitHub Copilot.
00:31:27.540 --> 00:31:29.540
So GitHub Copilot has a CLI.
00:31:29.960 --> 00:31:32.820
And there's a new GitHub Copilot app as well.
00:31:33.320 --> 00:31:37.020
And that app renders canvases in HTML as well.
00:31:37.540 --> 00:31:43.140
So I've seen some folks in my organization that are creating custom canvas apps for GitHub
00:31:43.140 --> 00:31:45.680
Copilot, the desktop app, because that's where they're living.
00:31:46.180 --> 00:31:49.000
And that's where they're doing their work out of kind of thing.
00:31:49.140 --> 00:31:54.640
So that's another super powerful concept is like kind of coaxing an LLM and getting it
00:31:54.640 --> 00:32:00.020
to build the view that you want and the experience you want in the place that you work.
00:32:00.280 --> 00:32:03.080
So going back to your skills, I have a lot of similar skills.
00:32:03.080 --> 00:32:08.420
Like I have a daily review skill that will go back and look at all my meeting notes from
00:32:08.420 --> 00:32:09.020
the day.
00:32:09.200 --> 00:32:10.680
It'll look at my calendar.
00:32:11.260 --> 00:32:15.240
Mine actually goes into, I don't know if you've tried this, it'll go into my team's
00:32:15.240 --> 00:32:15.840
conversations.
00:32:16.480 --> 00:32:22.560
And I've told it to look for certain, there's certain teams that I'm in that are super chatty
00:32:22.560 --> 00:32:25.760
where only like five or 10% of the stuff is relevant to me.
00:32:26.000 --> 00:32:32.480
And I'm like, go look through all of these chats and teams and give me, tell me where there's
00:32:32.480 --> 00:32:36.620
a chat related to a certain topic because that's the only thing I care about.
00:32:36.920 --> 00:32:41.540
So part of my review will actually analyze all my team's chats for me and tell me if there's
00:32:41.540 --> 00:32:45.120
anything I should go back and look at my team's chats or if I can just ignore all of them.
00:32:45.120 --> 00:32:46.040
I run a similar one.
00:32:46.380 --> 00:32:49.400
That's another automation or loop that I have out there running.
00:32:50.200 --> 00:32:53.500
So it looks across my inbox and my team's chats.
00:32:53.620 --> 00:32:54.080
And your team's.
00:32:54.080 --> 00:33:00.480
And it looks for emails from my manager, my skip, or people on my team.
00:33:00.540 --> 00:33:00.760
Okay.
00:33:00.900 --> 00:33:01.700
Who report to me.
00:33:02.560 --> 00:33:10.600
And initially that was a, let's get out there and kind of grab everything, right?
00:33:10.640 --> 00:33:13.700
Like the signal to noise ratio was, it was really bad.
00:33:13.700 --> 00:33:18.320
It was pretty horrible because like anytime somebody chatted with me on my team, like they
00:33:18.320 --> 00:33:20.960
could ping me on the side and say, how was the weather today?
00:33:21.200 --> 00:33:26.580
And I would all of a sudden get this alert that popped up and said, Joe is, Joe is, I
00:33:26.580 --> 00:33:30.280
don't have Joe on my team, but Joe is asking you how the weather is today.
00:33:30.380 --> 00:33:31.180
Do you want to respond?
00:33:31.520 --> 00:33:33.180
Oh no, that's, that's not important, right?
00:33:33.660 --> 00:33:36.100
So I definitely had to go and refine that thing in a similar way.
00:33:36.400 --> 00:33:39.380
I have a little bit of contextual filter, things like that.
00:33:39.380 --> 00:33:45.600
So it only alerts me now when folks are kind of pinging me with urgency or high priority
00:33:45.600 --> 00:33:51.100
items, particularly things that align to work streams that were kind of
00:33:51.100 --> 00:33:53.600
integrated on or that I need to know about what's happening.
00:33:54.100 --> 00:33:56.980
I've actually been thinking, and one of the things I would love to do is take it to the
00:33:56.980 --> 00:34:00.300
next level and just have it be able to respond back on teams.
00:34:00.920 --> 00:34:08.120
So like if somebody says, oh, hey, can you give me the latest update on feature ABC?
00:34:08.880 --> 00:34:13.380
If I already sent out the status update for feature ABC this week, I just wanted to send
00:34:13.380 --> 00:34:19.360
them the doc or send them a link like, or send them like the subject of that email that went
00:34:19.360 --> 00:34:20.300
out or whatever that was.
00:34:20.860 --> 00:34:26.340
And that feels like low hanging fruit and something that I could go out there and automate fullness
00:34:26.340 --> 00:34:26.900
of time.
00:34:27.260 --> 00:34:29.000
I just haven't been sure if it's worth it yet.
00:34:29.320 --> 00:34:33.520
Like putting the effort in to have an additional contextual filter, definitely worth it.
00:34:33.560 --> 00:34:37.300
Like I said, like I was getting all sorts of like stuff, like very much along the lines
00:34:37.300 --> 00:34:40.820
of like, so-and-so is telling you the weather's nice today.
00:34:41.020 --> 00:34:41.940
Like very important.
00:34:42.100 --> 00:34:42.740
Please respond.
00:34:43.100 --> 00:34:44.200
Like, no, I don't need to do that.
00:34:44.200 --> 00:34:45.020
I don't find mine too.
00:34:45.120 --> 00:34:46.360
It's just certain topics.
00:34:46.540 --> 00:34:51.220
But then in my dashboard, so I can run all of my skills from my dashboard.
00:34:51.580 --> 00:34:58.280
I can just click a button in the dashboard and it'll go open a bash prompt, a Claude session
00:34:58.280 --> 00:35:03.100
in the background and go run the skill in the background and I won't ever see it.
00:35:03.180 --> 00:35:07.920
So any skills that I don't have to interact with, then I have other skills where I know
00:35:07.920 --> 00:35:13.420
like I may go back and forth with them on or have more of a conversation with Claude.
00:35:13.640 --> 00:35:16.420
Those, the button actually copies a prompt to my clipboard.
00:35:16.520 --> 00:35:21.000
So I just open up a Claude session and paste the prompt right into a Claude session.
00:35:21.100 --> 00:35:22.720
Interesting that you chose to go down that path.
00:35:22.820 --> 00:35:24.140
Everything out of my dashboard.
00:35:24.300 --> 00:35:30.040
Go open the harness, go open Claude and run the skill versus me where I was like, I'm going
00:35:30.040 --> 00:35:34.020
to go, I'm going to go have it write the script and have it be super deterministic so I can
00:35:34.020 --> 00:35:38.440
run that script from a skill or from a button or from here.
00:35:38.540 --> 00:35:40.860
Some of my skills do still run scripts.
00:35:41.260 --> 00:35:42.300
It's the same.
00:35:42.300 --> 00:35:45.960
I still have a lot of scripts where those skills, it's still deterministic.
00:35:46.080 --> 00:35:49.960
It's a button, open a Claude session, run a script, run a skill.
00:35:49.960 --> 00:35:53.580
So one of the things maybe I was going at and something for folks to think about is, and
00:35:53.580 --> 00:35:58.780
one of the reasons I kind of go down this path quite a bit, is optimizing for token usage
00:35:58.780 --> 00:36:00.260
and context.
00:36:01.100 --> 00:36:07.040
So if you're using Claude as a harness, Claude has a fairly large system context already coming
00:36:07.040 --> 00:36:08.860
with it as part of the harness.
00:36:08.860 --> 00:36:13.280
And every time you're invoking a skill, you have to load that context no matter what.
00:36:13.280 --> 00:36:14.900
Like that's there.
00:36:14.900 --> 00:36:19.940
And so sure, you should have the advantages of cash tokens and things like that.
00:36:19.940 --> 00:36:24.500
If Anthropik's doing their job right, if they're not doing their job right on a given day, or
00:36:24.500 --> 00:36:29.000
stuff does go sideways, you're really in this place where you're burning tokens.
00:36:29.000 --> 00:36:33.240
And this stuff does cost money, right?
00:36:33.380 --> 00:36:37.500
Like, I'd be curious if you're comfortable saying like, do you have a sense for how much
00:36:37.500 --> 00:36:40.540
yours costs to run over the course of a month?
00:36:40.680 --> 00:36:41.040
If not?
00:36:41.120 --> 00:36:42.620
I do other things in it.
00:36:42.740 --> 00:36:47.580
And I mean, I will say I am paying for Claude team now.
00:36:47.840 --> 00:36:51.040
So I'm paying for the team edition of Claude.
00:36:51.160 --> 00:36:52.380
What's the team edition run?
00:36:52.560 --> 00:36:58.240
Which it's somewhere between, so they have like the Claude max, where it's the 5x or the
00:36:58.240 --> 00:36:58.840
20x.
00:36:58.960 --> 00:37:02.440
I'm actually sitting at like 6.5x.
00:37:02.440 --> 00:37:05.200
They don't have a 5x or 20x.
00:37:05.500 --> 00:37:08.680
So I would say I'm equivalent to the 5x, but it's a little bit more.
00:37:08.780 --> 00:37:11.120
I'm paying $125 a month for Claude.
00:37:11.200 --> 00:37:16.100
So my token consumption input and output is 100 to 120.
00:37:16.560 --> 00:37:20.720
I don't have the hard average in front of me, but I would say I probably average.
00:37:20.840 --> 00:37:26.420
I probably average like $110, $115 a month, something like that, to go through and run
00:37:26.420 --> 00:37:27.780
everything that I need to do.
00:37:28.340 --> 00:37:33.100
But that includes like the things like, hey, going and like bootstrapping the dashboard and
00:37:33.100 --> 00:37:33.660
things like that.
00:37:33.840 --> 00:37:37.980
I would expect like continued optimization as things come down.
00:37:38.120 --> 00:37:39.540
I was trying to find this the other day.
00:37:39.780 --> 00:37:44.060
Over the last 30 days, I can see things like product stickiness.
00:37:44.260 --> 00:37:45.000
This is connectors.
00:37:45.300 --> 00:37:46.680
I can see the skills.
00:37:47.620 --> 00:37:52.720
Most of these run in Claude code, usage, top skills, tool usage.
00:37:53.000 --> 00:37:59.580
I was just trying to look super quick if I could see how many gives me value.
00:38:00.280 --> 00:38:01.540
I haven't found.
00:38:02.100 --> 00:38:06.820
I'd have to dig through here like how many tokens I use.
00:38:07.040 --> 00:38:07.800
It's not free.
00:38:08.480 --> 00:38:09.500
No, it's not free.
00:38:09.560 --> 00:38:13.740
But I can fit it all in there with room to spare because I run these skills, but I also
00:38:13.740 --> 00:38:18.280
use it all the time for writing PowerShell scripts or analyzing data, that type of stuff.
00:38:18.620 --> 00:38:25.520
Because I do, similar to you, I have a lot of these skills that have turned into Python
00:38:25.520 --> 00:38:31.800
scripts in the background as well, where the skill calls a script to do a lot of the work
00:38:31.800 --> 00:38:32.160
for it.
00:38:32.280 --> 00:38:34.060
Python seems to be where it all lands.
00:38:34.280 --> 00:38:38.520
But a lot of what I do is very similar type of stuff that I built to you, except that
00:38:38.520 --> 00:38:41.180
I have multiple LLM wikis in that vault.
00:38:41.500 --> 00:38:43.760
And then I did the dashboard in Obsidian.
00:38:44.940 --> 00:38:49.500
It's been, at least from my end, super empowering because I get to work the way I want to work.
00:38:49.760 --> 00:38:55.180
I'm always frustrated at the friction of, here's Microsoft to do, and it does to do the way
00:38:55.180 --> 00:39:00.020
Microsoft wants to do them for that product or things like that.
00:39:00.360 --> 00:39:06.620
So I would love it if I could say, I could take this thing and package it up and give it
00:39:06.620 --> 00:39:08.480
to you and it would work for the way you work.
00:39:08.720 --> 00:39:11.600
I don't think that's the case, nor do I think it's the point.
00:39:11.820 --> 00:39:18.600
The whole point is to have the way you work and the set of skills and loops and sub-agents
00:39:18.600 --> 00:39:25.540
or authentic flows and connectors and everything else that are out there to wire everything
00:39:25.540 --> 00:39:28.540
up and get it going and have a good set of outcomes for it.
00:39:28.680 --> 00:39:32.900
I really do feel like it helps accelerate the work that I do.
00:39:33.440 --> 00:39:35.880
I think it keeps me better grounded.
00:39:36.580 --> 00:39:39.980
I still have to go through a lot of context switching and I find it super helpful to have
00:39:39.980 --> 00:39:43.260
a lot of the grounded notes and things just right there at my fingertips.
00:39:43.260 --> 00:39:49.420
Go ask a question so I can search in Obsidian or when I do need to use an LLM to iterate
00:39:49.420 --> 00:39:51.020
over it, great, here you go.
00:39:51.120 --> 00:39:58.420
I can hop in that directory and invoke my harness of choice and I'm off to the races.
00:39:58.680 --> 00:40:02.840
I did things a little differently too for my tasks because I have contractors working for
00:40:02.840 --> 00:40:03.040
me.
00:40:03.360 --> 00:40:08.640
So I want to make sure tasks are updated in Asana and some of that.
00:40:08.640 --> 00:40:14.940
So my dashboard, instead of managing tasks here, all of my skills and prompts and everything
00:40:14.940 --> 00:40:17.740
ensure that Asana stays the source of truth.
00:40:18.080 --> 00:40:23.920
So if I see something on my list of my top to do or something that I need to get done, it
00:40:23.920 --> 00:40:29.340
actually links to Asana so I can click it, open it in Asana and I check it off there and
00:40:29.340 --> 00:40:31.420
then the dashboard will update.
00:40:31.980 --> 00:40:33.780
And same thing for on fire.
00:40:33.900 --> 00:40:36.900
I have two right now where there's an email and a Teams conversation.
00:40:37.080 --> 00:40:42.660
I have created my dashboard so I can click on those items that I've pulled in and it takes
00:40:42.660 --> 00:40:46.460
me right to the email or it takes me right to the Teams conversation.
00:40:46.840 --> 00:40:53.200
I have one for blog posts that takes me right to the draft blog posts or to the client record
00:40:53.200 --> 00:40:54.160
in my CRM.
00:40:54.280 --> 00:41:00.420
So my dashboard, the way I've thought about mine, I've turned my dashboard and my LLM into
00:41:00.420 --> 00:41:07.060
an integration point so I can pull all the information from all the systems I want and built a central,
00:41:07.080 --> 00:41:12.760
dashboard command center that I can work from but then link out to all the different
00:41:12.760 --> 00:41:15.720
things that I need to respond to that pop up on it.
00:41:15.800 --> 00:41:17.540
I'm on a very similar path.
00:41:17.740 --> 00:41:22.220
I have this kind of backlog of things that I want to go kind of integrate next.
00:41:22.460 --> 00:41:28.760
So one is, I mentioned we use Azure DevOps, ADO, for kind of tracking of epics and features
00:41:28.760 --> 00:41:30.420
and PBIs and things like that.
00:41:30.420 --> 00:41:36.340
I'd love to have that integrated view right there for, hey, I'm working on project XYZ,
00:41:36.720 --> 00:41:42.340
click on project XYZ, here's the ADO items for it, and here's their current state and status,
00:41:42.980 --> 00:41:44.460
things like that.
00:41:44.740 --> 00:41:50.260
The other one that I want to build is, at some point, I struggle with this stuff, I don't
00:41:50.260 --> 00:41:51.460
know when I'm reinventing the wheel.
00:41:52.260 --> 00:41:56.980
This one feels a little bit maybe like reinventing SharePoint, but I've got these projects that
00:41:56.980 --> 00:41:57.300
exist.
00:41:57.680 --> 00:42:02.720
So when you do the Carpathia-style LLM wikis, you end up with artifacts and entities, you end
00:42:02.720 --> 00:42:07.720
up with these discrete templates for types of objects in your LLM wiki.
00:42:08.180 --> 00:42:13.320
So I've got projects and work streams and events and things like that.
00:42:13.920 --> 00:42:18.880
So for my work streams and my projects, there's these kind of canonical briefs that sit there
00:42:18.880 --> 00:42:21.240
that have all the detail about them.
00:42:21.760 --> 00:42:24.800
And then they have all the wiki links to sublinks and things like that.
00:42:25.240 --> 00:42:30.440
But sometimes the sublinks they carry are links to documents, so like Word docs, Excels,
00:42:30.620 --> 00:42:31.220
PowerPoints.
00:42:31.220 --> 00:42:39.680
So I kind of want to view where I can say, hey, for Workstream ABC, or Workstream A, click
00:42:39.680 --> 00:42:42.480
on Workstream A, here's all the important documents for it.
00:42:42.580 --> 00:42:46.340
Like, here's all the contextual things and just like the little snippet and summary of
00:42:46.340 --> 00:42:46.820
it that's there.
00:42:47.100 --> 00:42:50.680
And then like the document, to be very clear, like the artifacts, the canonical artifact,
00:42:50.900 --> 00:42:51.920
not stored in my wiki.
00:42:52.140 --> 00:42:56.220
Like that stuff's still IRM, it's got sensitivity labels, all that stuff.
00:42:56.440 --> 00:42:58.220
But the link to it is in my wiki.
00:42:58.220 --> 00:43:02.840
So now it's like one-stop shopping for me, where in the past, I would have been going
00:43:02.840 --> 00:43:08.360
to SharePoint and just going to kind of like the generic intranet search or even like Copilot
00:43:08.360 --> 00:43:08.860
search.
00:43:09.020 --> 00:43:12.740
I'm finding it to be kind of annoying because it's an LLM-based search.
00:43:12.920 --> 00:43:16.800
So you kind of start talking to it like an LLM and it goes sideways pretty quick.
00:43:17.520 --> 00:43:19.040
So I've got all this stuff sitting.
00:43:19.340 --> 00:43:20.480
Like the context is there.
00:43:20.640 --> 00:43:25.320
It's just about, like you said, integrating the views and kind of making it one-stop shopping
00:43:25.320 --> 00:43:27.100
across all these various systems.
00:43:27.100 --> 00:43:32.940
Yeah, my project for my dashboard has a backlog markdown file in it where I just, when I come
00:43:32.940 --> 00:43:35.160
up with an idea, I throw an idea in there.
00:43:35.340 --> 00:43:38.760
So when I have some time or I'm sitting on the couch in the evening and I want to build
00:43:38.760 --> 00:43:42.880
a new feature on my dashboard, I have my whole backlog and I can pick one off the backlog
00:43:42.880 --> 00:43:43.720
to start building.
00:43:43.840 --> 00:43:44.640
Anything we miss.
00:43:44.860 --> 00:43:47.900
I'm sure this is one we're going to come back to because I feel like, to your point,
00:43:47.980 --> 00:43:52.940
you're always evolving it, changing it, modifying it as everything changes.
00:43:53.240 --> 00:43:54.800
And as you come up with ideas.
00:43:54.800 --> 00:43:58.400
The builder's itch comes in along the way.
00:43:58.800 --> 00:44:03.980
And then the other thing I like about it is, so you not only get to go build these things,
00:44:04.280 --> 00:44:07.220
but like I said, you get to adapt them to the way you work.
00:44:07.700 --> 00:44:11.500
And you also get to adapt them to your taste.
00:44:12.040 --> 00:44:19.000
Like I said, I don't have the greatest reaction sometimes when I have to go into a Microsoft
00:44:19.000 --> 00:44:24.420
to-do or a planner, but it doesn't rub me nearly as much the wrong way going into this
00:44:24.420 --> 00:44:29.060
thing that I built for the way I guided an LM to build it for me, because there's no way
00:44:29.060 --> 00:44:32.660
I was writing all that stuff myself to go and put it together.
00:44:32.660 --> 00:44:36.020
So it actually, it's like, it's a little bit more calming, right?
00:44:36.100 --> 00:44:40.300
Like while it's bringing in probably, like I said, more stuff.
00:44:40.520 --> 00:44:46.480
And if anything, like it's allowing me to over-function, but it's at least allowing me to over-function
00:44:46.480 --> 00:44:51.880
in like my way of over-functioning without like my head constantly on a swivel.
00:44:52.080 --> 00:44:56.900
That's the biggest advantage I've found of this and frustrations I've had other times
00:44:56.900 --> 00:45:01.820
that I'm like, how do I integrate Asana with Teams, with Outlook, in a way that works for
00:45:01.820 --> 00:45:02.820
how I want to work?
00:45:03.180 --> 00:45:09.940
This has allowed me to pull that information and put some integrations in place that A,
00:45:09.980 --> 00:45:12.380
I can never develop or build on my own.
00:45:12.780 --> 00:45:18.060
And B, I can use products, but then tie them all together in a way that makes sense for
00:45:18.060 --> 00:45:22.000
the way I want to see information, work with information, manage my day.
00:45:22.080 --> 00:45:22.700
Et cetera.
00:45:23.100 --> 00:45:27.120
So that's the part I've enjoyed about this, to your point.
00:45:27.380 --> 00:45:29.260
It's, I can tweak it to me.
00:45:29.600 --> 00:45:31.260
So with that, just wrap it up.
00:45:31.340 --> 00:45:32.840
We went almost an hour today, Scott.
00:45:33.040 --> 00:45:34.020
Sometimes we go a little bit longer.
00:45:34.600 --> 00:45:37.140
So I'm always interested in feedback from other folks.
00:45:37.560 --> 00:45:41.900
We've mentioned a couple of times and it's good to see people coming and commenting on
00:45:41.900 --> 00:45:43.140
Reddit or LinkedIn.
00:45:43.720 --> 00:45:46.680
I would love to hear what other folks are doing there.
00:45:46.680 --> 00:45:52.780
Like, what's your superpower, your unlock for LLM wikis?
00:45:53.020 --> 00:45:54.340
Like, are you using them?
00:45:54.760 --> 00:45:59.780
I think you and I are maybe a little bit more on like one end of the spectrum where we're
00:45:59.780 --> 00:46:04.420
like in CLI harnesses or a little bit more advanced thing.
00:46:04.560 --> 00:46:11.020
Like, I've seen folks on definitely like YouTube using very much like Claude Cowork style kinds
00:46:11.020 --> 00:46:14.280
of things to go and execute these same workflows.
00:46:15.320 --> 00:46:20.720
So I'd love to understand where others are experimenting, what they're doing, where they're finding the
00:46:20.720 --> 00:46:25.540
unlock, and maybe what else we're missing that could be like the take it up another notch.
00:46:25.960 --> 00:46:27.380
We'll do the final tap thing.
00:46:27.540 --> 00:46:28.460
Like, does this go to 11?
00:46:28.680 --> 00:46:29.620
Yes, yes, it does.
00:46:29.760 --> 00:46:30.000
Yes.
00:46:30.160 --> 00:46:30.740
Let us know.
00:46:31.000 --> 00:46:31.700
We'd love to hear it.
00:46:31.940 --> 00:46:32.220
All right.
00:46:32.420 --> 00:46:33.060
Well, that's Scott.
00:46:33.260 --> 00:46:34.160
It's time for the weekend.
00:46:34.520 --> 00:46:36.000
Almost time for the weekend.
00:46:36.420 --> 00:46:37.060
But yeah.
00:46:37.620 --> 00:46:39.040
Well, as always, thanks, Ben.
00:46:39.300 --> 00:46:39.740
Not a problem.
00:46:40.000 --> 00:46:40.400
Thank you.
00:46:40.540 --> 00:46:41.380
And we'll talk to you again soon.
00:46:43.480 --> 00:46:47.300
If you enjoyed the podcast, go leave us a five star rating in iTunes.
00:46:47.300 --> 00:46:52.420
It helps to get the word out so more IT pros can learn about Office 365 and Azure.
00:46:53.340 --> 00:46:57.460
If you have any questions you want us to address on the show or feedback about the show,
00:46:57.900 --> 00:47:01.400
feel free to reach out via our website, Twitter, or Facebook.
00:47:01.860 --> 00:47:03.920
Thanks again for listening and have a great day.
00:00:00.000 --> 00:00:11.420
Welcome to Episode 438 of the Microsoft Cloud IT Pro Podcast, recorded live on October 2, 2026.
00:00:11.960 --> 00:00:21.280
This is a show about Microsoft 365 and Azure from the perspective of IT pros and end users, where we discuss a topic, more recent news, and how it relates to you.
00:00:21.280 --> 00:00:32.520
Back in Episode 429, Scott and Ben introduced LLM wikis. Today, they're circling back around to see what they've turned into after using them daily for a few months.
00:00:32.860 --> 00:00:42.820
Scott walks through the skills that now run his daily life, from a morning prioritization pass and meeting prep to meeting capture and end-to-end daily reviews.
00:00:42.820 --> 00:00:50.660
We also talk about growing a workflow from an on-demand skill to a scheduled skill, loops, and deterministic scripts.
00:00:51.020 --> 00:00:59.860
We also discuss different approaches Scott and Ben have to creating custom dashboards and command centers and how they interact with their data.
00:01:01.880 --> 00:01:07.460
We're live with another episode of the Microsoft Cloud IT Pro Podcast, live from our houses.
00:01:07.740 --> 00:01:09.380
Live from New York. It's Saturday night.
00:01:09.640 --> 00:01:11.160
Oh, no, that's a different one.
00:01:11.280 --> 00:01:13.700
Different show. We're not quite that popular, Scott.
00:01:13.840 --> 00:01:14.160
Someday.
00:01:14.160 --> 00:01:20.480
I know we look that good, and we sound that good, and we could clearly do a primetime, not really primetime, late night,
00:01:20.760 --> 00:01:24.580
NBC show. Better be careful. We're going to end up talking about copyrighted stuff.
00:01:24.640 --> 00:01:28.260
We could go see if we could end up on The Onion and the new InfoWars channel or something.
00:01:28.440 --> 00:01:34.440
We could just spin yarn and conspiracy theories about Azure and Microsoft 365.
00:01:34.740 --> 00:01:38.100
365. We could. We could come up with all kinds of conspiracy theories.
00:01:38.200 --> 00:01:40.280
My daughter is actually enjoying that right now.
00:01:40.340 --> 00:01:40.980
Do you know it'll be fine?
00:01:40.980 --> 00:01:50.920
Yes, my daughter's actually taking a class right now that's all about conspiracy theories, where they spend like a week, because she's homeschooled.
00:01:50.920 --> 00:02:01.320
So all the kids in the class go home from their co-op, spend a week like looking at different conspiracy theories, and then come back to class the next week and debate the various conspiracy theories.
00:02:01.720 --> 00:02:09.960
So it's kind of interesting. She's having fun. They did the moon landing today. So the last week, she's been looking at all the different conspiracy theories around the moon landing.
00:02:09.960 --> 00:02:25.600
My oldest son is, he's studying history in college. And so he's, so he's in his second year now when he's getting kind of like more into things. And he has had the realization that history repeats itself.
00:02:27.000 --> 00:02:32.300
You didn't learn this in like middle school or high school, but you're coming to it in college. Okay, well, good realization.
00:02:32.680 --> 00:02:35.620
I mean, at least he came to that realization.
00:02:35.620 --> 00:02:40.120
There's some maturity there. I have hope that he'll move out of my house someday. We'll see.
00:02:40.880 --> 00:03:05.540
So you know what else happened today, Scott? I called my HVAC company because we have like in Florida, AC is kind of important. So twice a year, we have like regular, we have a plan. We sign up for discounted maintenance and all that. But then they do like twice annual, twice annual maintenance checks. And I got a text message and an email and all of it. Hey, it's time to schedule one.
00:03:05.540 --> 00:03:23.060
So I called them today because usually that works better than trying to do it online and all of that. And it was like, you're on hold. And then I had someone picked up the phone and it was like, I can't remember the name. Hi, I'm Lisa. This is your AI assistant. Like it was an AI.
00:03:23.060 --> 00:03:35.260
Well, he should announce itself that way. Person. I'd rather fall into like a phone tree and interactive voice recording like that, where it's very upfront and clear that yes.
00:03:35.260 --> 00:03:46.300
Yeah. And that's what it did. Name. But it was a little, it was interesting. I wanted to be annoyed at it the whole time because I knew I was talking to AI, but it actually went really smooth.
00:03:46.460 --> 00:04:00.360
They superimposed normal office sounds in it. So I could hear like a keyboard clicking in the background and all the stuff. But this AI assistant was, why are you calling? I was like annual maintenance and asked me all the questions it needed.
00:04:00.360 --> 00:04:14.920
Like had an actual conversation with this AI agent, just like I would have had with anybody in their office and was able to schedule my appointment, got the text message and the email with the confirmation of the appointment date and time.
00:04:14.920 --> 00:04:28.020
And as much as I wanted to be annoyed knowing I was talking to AI, it actually was a good experience. And I was pleasantly surprised and impressed with how well and smoothly it went.
00:04:28.020 --> 00:04:35.940
Good to hear. When you were texting me about it and telling me about it, I thought you were going to tell me like you fired off because there's lots of personal agents out there right now.
00:04:36.120 --> 00:04:45.540
So you've got Meadow with Muse, there's some invite stuff like Instinct that's out there. OpenAI announced Dots at their dev day this week.
00:04:45.960 --> 00:04:55.200
So I thought what you were going to tell me was like, oh, you had your AI personal assistant, because now these things are starting to come with phone numbers, and do inbound outbound calling.
00:04:55.200 --> 00:05:04.680
You had your AI assistant call another AI assistant, and maybe they got caught in one of those loops of like talking to each other in like binary or like whatever.
00:05:05.300 --> 00:05:06.440
No, but that would be funny.
00:05:06.780 --> 00:05:12.060
Sound and wavelengths. They wanted to have like a waveform representation to go and talk back and forth.
00:05:12.220 --> 00:05:16.500
I should try that next time is have my bot call their bot.
00:05:16.720 --> 00:05:25.100
Speaking of bots and Dots, did you also see the somewhat trolling, hilarious news like OpenAI announced their Dots?
00:05:25.100 --> 00:05:31.140
And immediately after that, Musk went out and bought Dot.com and redirected it to the GrokBot.
00:05:34.480 --> 00:05:38.740
If you want a domain name about a product, you're going to buy the domain name before you announce it.
00:05:38.820 --> 00:05:43.080
Or you just get big enough that you have your own company domain name, right?
00:05:43.160 --> 00:05:46.160
Like the whole like Cloud, Dot, Microsoft, and things like that.
00:05:46.720 --> 00:05:47.860
See, that would solve it too.
00:05:47.960 --> 00:05:49.800
That's the other way to go about it.
00:05:49.920 --> 00:05:52.600
But yeah, we should talk about AI stuff today.
00:05:52.760 --> 00:05:53.240
We should.
00:05:53.240 --> 00:06:09.540
I want to revisit the talk that we had a little while ago about getting started with LL and Wikis and kind of talking about what they can turn into with a little bit of maturity.
00:06:10.080 --> 00:06:14.920
So like this will be a little bit more of like a freeform conversation for us.
00:06:15.100 --> 00:06:16.760
I want to talk about my setup.
00:06:17.060 --> 00:06:20.040
I want to learn a little bit more about your setup.
00:06:20.640 --> 00:06:28.820
And I really want to dig into kind of the workflows and maybe like the skills and agents that you're using.
00:06:28.820 --> 00:06:39.600
Like what are the harnesses that you're using around this that make it helpful for you and empower your day-to-day work, accelerate your day-to-day work.
00:06:39.940 --> 00:06:41.160
Maybe they make things harder, right?
00:06:41.160 --> 00:06:42.840
As you've been experimenting and going through.
00:06:43.100 --> 00:06:46.980
So what are things that maybe you hit a brick wall with or I hit a brick wall with?
00:06:47.320 --> 00:06:52.060
And we kind of went in a little bit of a different direction.
00:06:52.620 --> 00:06:54.580
So yeah, you want me to go first?
00:06:54.680 --> 00:06:55.040
Sure.
00:06:55.160 --> 00:06:55.640
What I've been doing?
00:06:55.780 --> 00:06:56.060
Yes.
00:06:56.200 --> 00:06:57.640
I'm curious to hear what you've been doing.
00:06:58.100 --> 00:07:02.420
So I don't want to get to your point earlier about like copyright infringement, things like that.
00:07:02.580 --> 00:07:03.620
I don't want anything there.
00:07:03.620 --> 00:07:17.220
But I took what were like various instantiations of LL and Wikis that I've been working with around like various areas of like research for specific projects and initiatives that I work on.
00:07:17.320 --> 00:07:20.280
Just like my day-to-day tasks, tracking, things like that.
00:07:20.660 --> 00:07:27.520
And I consolidated them all into one wiki and it's my WorkOS, right?
00:07:27.900 --> 00:07:29.420
It's my work operating system.
00:07:30.040 --> 00:07:37.140
So it's meant to track really everything that I do across work, whether that's working.
00:07:37.560 --> 00:07:41.460
So I'm a product manager, so that could be working on a new product.
00:07:41.740 --> 00:07:43.940
It could be working on an update to an existing feature.
00:07:44.180 --> 00:07:46.160
It could be like building a new feature.
00:07:46.320 --> 00:07:47.980
Maybe that's not a full product, right?
00:07:47.980 --> 00:07:51.560
But it's complimentary to existing product suite, things like that.
00:07:51.560 --> 00:08:08.140
And I do all the PME things that come along with that of market research, writing PRDs, specs, having customer conversations, wrangling work management systems like ADO.
00:08:08.560 --> 00:08:09.640
We use Azure DevOps.
00:08:10.380 --> 00:08:13.040
Maybe somebody's using JIRA, something like that.
00:08:13.040 --> 00:08:23.580
But really, everything that I do gets logged, gets tagged, and goes into that WorkOS.
00:08:24.300 --> 00:08:28.700
So off the cuff, that sounds like, okay, but how do I get everything in there?
00:08:29.280 --> 00:08:33.140
And what's important and what happens along the way?
00:08:33.780 --> 00:08:44.080
So I've got a set of skills that are out there that I've put together that work for my way of work and that operate within the harness that I use.
00:08:44.340 --> 00:08:50.380
So the harness that I use today is Microsoft Scout, the desktop app for Scout.
00:08:50.740 --> 00:08:55.380
I'm going to be a little bit sad when that goes to autopilot here in the future because I don't know what that's going to do.
00:08:55.560 --> 00:08:57.820
Not autopilot because we have Microsoft autopilot.
00:08:58.000 --> 00:09:01.400
We need to differentiate now, Scott, because we have Microsoft autopilot.
00:09:01.560 --> 00:09:02.880
Autopilot inside of copilot.
00:09:03.000 --> 00:09:03.800
Copilot autopilot.
00:09:03.900 --> 00:09:05.180
And copilot autopilot.
00:09:05.520 --> 00:09:07.540
That's another topic for another day.
00:09:07.660 --> 00:09:08.680
We'll get to that one eventually.
00:09:09.400 --> 00:09:23.780
So I've got this set of skills that help me with inputs and outputs to get information into that wiki and then help me get good concrete outputs out of it.
00:09:23.780 --> 00:09:33.760
So one would be, for example, one of the ones I use quite a bit actually is an email thread ingester skill.
00:09:34.380 --> 00:09:41.680
So my work takes place across teams, email, in-person conversations, teams meetings.
00:09:41.680 --> 00:09:44.700
If I'm meeting with customers, it could be Zoom meetings, things like that.
00:09:45.200 --> 00:09:53.000
But particularly for internal conversations, a lot of that stuff still happens in email or there's newsletters and various things that go out.
00:09:53.660 --> 00:09:57.960
So I have a little skill that's my WorkOS email thread ingester.
00:09:59.040 --> 00:10:09.320
And because it works in context of WorkIQ and the Microsoft Graph, it respects sensitivity labels, DLP, all those kinds of things.
00:10:09.320 --> 00:10:17.920
It's not like I'm taking super secret stuff out of the system and throwing it over in this random text wiki over here that lives in my OneDrive or anything like that.
00:10:18.300 --> 00:10:19.520
So I use that one quite a bit.
00:10:19.880 --> 00:10:24.420
And that's super handy and helpful for me for point-in-time kinds of things.
00:10:25.160 --> 00:10:32.960
I have a skill that runs every single morning for me that goes every single weekday for me.
00:10:32.960 --> 00:10:41.420
It runs every morning at 9 a.m. local, which is my WorkOS daily prioritization skill.
00:10:42.220 --> 00:10:48.000
So what that does is it looks at the to-dos that are in my WorkOS.
00:10:48.340 --> 00:10:55.080
So I made this very conscientious decision not to use Planner or Microsoft to-do or anything like that.
00:10:55.220 --> 00:10:59.320
I'm just kind of managing everything in markdown files inside my WorkOS.
00:10:59.620 --> 00:11:04.360
So your source of truth for your tasks is these markdown files in WorkOS?
00:11:04.640 --> 00:11:04.900
Yes.
00:11:05.100 --> 00:11:05.320
Okay.
00:11:05.420 --> 00:11:08.080
But I find that works because I can tag them.
00:11:08.500 --> 00:11:10.220
They carry the tags through the wiki.
00:11:10.620 --> 00:11:13.720
They carry all the grounding and the context.
00:11:14.000 --> 00:11:22.400
I can link to-dos very easily to projects or to initiatives and various other contexts that exist in that wiki.
00:11:22.640 --> 00:11:38.040
So what my kind of daily prioritization skill does for me is it's meant to look at my prior day and or my incomplete tasks that are sitting out there that I've already spec'd out and help me kind of rationalize,
00:11:38.440 --> 00:11:51.800
here's the number one thing that you should do today out of your existing to-do list based on what happened between yesterday and today through your email, through your team signals, all those kinds of things, documents people shared with you.
00:11:52.220 --> 00:11:55.080
Here's another task that you should go and do.
00:11:55.580 --> 00:12:04.900
And then if it has room for it, the way I structured the skill is I always want to have kind of like three things on my plate in front of me that are like these are like the must-dos for today kind of thing.
00:12:05.420 --> 00:12:06.280
So that sits out there.
00:12:06.280 --> 00:12:09.400
I have a question about that one because I have one similar to that.
00:12:09.740 --> 00:12:30.020
Have you, and this is something I've done with mine with some success, have you had that skill when it looks at your daily to-dos, also look at your meetings for the upcoming day to prioritize if there's anything that you have to have done for a meeting coming up in like the next eight hours based on your tasks and projects as well?
00:12:30.200 --> 00:12:30.780
Yes and no.
00:12:30.940 --> 00:12:33.440
So we should get into that one next.
00:12:33.520 --> 00:12:33.740
All right.
00:12:33.740 --> 00:12:40.580
I want to kind of start the conversation here talking about the skills that I've integrated and that I'm using within Scout as a harness.
00:12:40.860 --> 00:12:45.040
I could be using Copilot, CLI, Claude, Codex, Anti-Gravity.
00:12:45.140 --> 00:12:45.900
It really doesn't matter, right?
00:12:46.000 --> 00:12:51.680
A skill is a skill and that markdown is portable and can come over and I can tie my connectors, all that kind of stuff.
00:12:52.200 --> 00:13:02.320
But there's this concept that I think becomes very powerful of moving from skills to loops and having loops that go out and execute things for me.
00:13:02.680 --> 00:13:07.320
So while like my daily prioritization skill is just a skill, I could run it on demand.
00:13:07.560 --> 00:13:09.660
I could come in and I could say, hey, run this thing.
00:13:10.080 --> 00:13:12.820
It's both a skill and it's a loop for me.
00:13:13.180 --> 00:13:18.680
So I put it in the harness and I schedule it and I say, hey, run this every day at 9 a.m.
00:13:18.680 --> 00:13:30.260
I actually have, I can run the same skill like every 30 minutes basically on a constant loop and have it do constant reevaluation for me like throughout the day and pivot things through.
00:13:30.260 --> 00:13:44.620
So I have two other skills when it comes to meetings that also function as skills and or more harnessed loops and kind of just on the flywheel of they're going to do things automatically for me.
00:13:44.980 --> 00:13:48.160
So one is a meeting prep skill.
00:13:48.540 --> 00:13:54.340
So I have like WorkOS-MeetingPrep and I have WorkOS-MeetingCapture.
00:13:54.600 --> 00:13:57.820
So WorkOS-MeetingPrep is my prep skill.
00:13:57.820 --> 00:14:16.640
So as my daily prioritization skill is running in the morning and it's going through like that harness and that automated scheduling and now it's turning more into a loop, what that's actually doing, one of the subtasks that it has is it goes and runs meeting prep for every single meeting that I have that day.
00:14:16.700 --> 00:14:18.100
That's on my calendar, okay?
00:14:18.680 --> 00:14:27.040
And then as I'm coming out of meetings, many of the meetings we attend are recorded, they have transcripts, they have all of those things.
00:14:27.040 --> 00:14:28.700
I have a meeting capture skill.
00:14:29.240 --> 00:14:38.520
And my meeting capture skill goes and evaluates the recording, it evaluates the transcript, it grabs the WorkIQ context, which is inclusive of like the meeting chat.
00:14:38.760 --> 00:14:42.160
So you can catch like documents that were shared in the meeting, things like that.
00:14:42.660 --> 00:14:50.500
And that comes through and effectively it turns my meetings into a two-step flow of prep and then capture.
00:14:50.860 --> 00:14:58.340
Or if I have a meeting that maybe, let's say like I was an optional attendee or maybe it was a learning forum, it's something that I want to catch up on later.
00:14:58.780 --> 00:15:00.000
So I wasn't active in it.
00:15:00.300 --> 00:15:09.020
For like meetings on my calendar that I'm just following or I'm an optional attendee, things like that, my morning prioritization skill doesn't actually run prep for those automatically.
00:15:09.140 --> 00:15:10.060
It just skips over them, right?
00:15:10.060 --> 00:15:11.340
I don't need them right now.
00:15:11.740 --> 00:15:21.900
So for those, if I want to, I can go in and I can basically run, I kind of have this like overarching maybe superset skill, which is generate and capture.
00:15:22.280 --> 00:15:29.160
So I can run generate individually, I can run capture post-meeting, or I can run generate and capture, which kind of does both for me.
00:15:29.360 --> 00:15:36.840
And generating capture is helpful for the meeting that's already happened that I didn't attend because they do different things, right?
00:15:36.940 --> 00:15:43.860
Prep goes out and looks at my inbox, it looks at conversations with meeting attendees, and it says, hey, what's the context for this thing?
00:15:44.280 --> 00:15:54.580
Capture is go blend that kind of meeting brief and the meeting prep with the actual outcomes of that meeting and what's next along the way.
00:15:55.240 --> 00:16:04.340
And then I have another skill that runs in a loop, again, on an automation, which is my end-of-day wrap-up.
00:16:04.640 --> 00:16:12.260
And part of my end-of-day wrap-up is that skill evaluates my wiki and it says, what are the meetings I attended today that I forgot to capture?
00:16:12.260 --> 00:16:19.560
Or I forgot to do a prep and capture on, and it just goes and grabs them automatically for me and kind of brings that context in.
00:16:19.960 --> 00:16:27.240
And that's good for me because then in that virtuous cycle of tomorrow's a new day, my automation for my daily prep runs again.
00:16:27.700 --> 00:16:31.080
And now it's able to catch up on the meetings and the context and things that are there.
00:16:31.420 --> 00:16:40.160
So I don't have to constantly be like human intervention in the loop to say, like, click a button or go run this skill, go run this skill, go run this skill in my harness.
00:16:40.440 --> 00:16:48.560
It's just happening kind of automatically and automagically as I go through my day.
00:16:49.240 --> 00:17:00.160
And then, let's see, because this is an LLM wiki, like, if people are on, like, the Carpathi-style LLM wiki path, they often end up with, like, that raw or clippings or ingest folder for things.
00:17:00.540 --> 00:17:01.620
Well, I'm a PM.
00:17:01.760 --> 00:17:02.400
I do research.
00:17:02.520 --> 00:17:04.380
I have to keep up on industry news.
00:17:04.740 --> 00:17:08.500
It could be Gartner, IDC, launch from a competitor, things like that.
00:17:08.760 --> 00:17:12.580
So I'm just running things in a markdown wiki at Subsidion.
00:17:12.800 --> 00:17:14.560
So I can use Subsidion Web Clipper.
00:17:14.980 --> 00:17:18.340
So I send things into a, like, a raw or a clippings directory.
00:17:18.880 --> 00:17:21.600
So I have a, hey, go ingest my clippings.
00:17:21.880 --> 00:17:23.460
Like, go rationalize over this.
00:17:23.820 --> 00:17:26.120
And again, organize it in context of my projects.
00:17:26.120 --> 00:17:26.740
Tag it.
00:17:27.500 --> 00:17:31.060
Do all those things that you need to do along the way.
00:17:31.700 --> 00:17:34.120
And then I have these kind of, I don't run them in loops.
00:17:34.300 --> 00:17:35.060
I don't run them automatically.
00:17:35.280 --> 00:17:39.800
I have a bunch of, I call them, like, my little, like, sub-agents or my helper skills.
00:17:40.700 --> 00:17:44.760
So I'm responsible for a portion of a business.
00:17:44.940 --> 00:17:48.060
That business has, like, a monthly business review.
00:17:48.580 --> 00:17:50.300
We do ongoing monthly product reviews.
00:17:50.600 --> 00:17:53.520
We do weekly execution reviews, things like that.
00:17:54.040 --> 00:18:00.480
And those often require, like, very structured inputs of, let's go grab your ADO items, get the current status.
00:18:00.480 --> 00:18:03.520
What's the dates, what's the update, things like that.
00:18:04.100 --> 00:18:12.920
So I have these skills that go out and help me do those things, like WorkOS MBR prep, WorkOS product review prep, things like that.
00:18:13.480 --> 00:18:21.120
And they all go out and they kind of work and they help me get the rough outlines for those things.
00:18:21.380 --> 00:18:23.480
I don't let them, like, write everything for me.
00:18:23.820 --> 00:18:28.100
I have this very visceral reaction to, like, just pure, like, AI-written things.
00:18:28.280 --> 00:18:37.840
I'll let them bullet it out so that then I can take them and kind of put my own thinking or my own spin on top of it along the way.
00:18:38.160 --> 00:18:47.100
And then the last thing that I did, and this is probably, like, super niche, but it's been very effective for me,
00:18:47.340 --> 00:18:52.560
is I took the things that I was doing in the harness.
00:18:53.380 --> 00:18:55.620
So I started in Copilot CLI, eventually moved to Scout.
00:18:55.820 --> 00:18:57.280
Now Scout's kind of like sitting over there.
00:18:57.620 --> 00:18:59.620
So Scout's really like my automation engine.
00:18:59.940 --> 00:19:02.280
It's not actually where I go and run these things from.
00:19:02.640 --> 00:19:09.020
So if I need to go, say, run a meeting prep, so I've got an upcoming meeting and that prep hasn't been run yet,
00:19:09.100 --> 00:19:13.520
or I've got to go run a capture, I built myself a WorkOS dashboard.
00:19:14.020 --> 00:19:24.380
So I have a little local HTML site that's served to me, and that site presents to me my calendar.
00:19:24.960 --> 00:19:29.360
It presents the meetings that haven't been prepped versus those that have been prepped,
00:19:29.480 --> 00:19:31.940
those that haven't been captured versus those that haven't been captured.
00:19:32.380 --> 00:19:34.280
It gets my to-dos there for me.
00:19:34.280 --> 00:19:40.720
So rather than me having to go into Obsidian, say I'm updating a to-do from not started to in progress,
00:19:41.180 --> 00:19:44.420
I actually, I go to this website, and I just click the to-do.
00:19:44.580 --> 00:19:47.060
It's kind of like to-doist or something for me just sitting there.
00:19:47.200 --> 00:19:51.680
And I update it right there, and that thing goes and writes the wiki, the markdown for me.
00:19:51.760 --> 00:19:54.340
So I don't have to go, like, find the markdown file and worry about, like,
00:19:54.340 --> 00:19:55.800
oh, did I get the tag right, everything.
00:19:56.300 --> 00:19:58.980
Things are very deterministic for me.
00:19:59.600 --> 00:20:03.420
So I've got the wiki there when I need it, when I need to go search for it.
00:20:03.480 --> 00:20:04.560
I do all that through Obsidian.
00:20:04.840 --> 00:20:09.040
I've got Scout sitting on the side, which is running my loops and running my automations
00:20:09.040 --> 00:20:13.740
and is kind of the house for those skills in the execution engine.
00:20:14.140 --> 00:20:17.740
And then I've got this just kind of dashboard that I live in.
00:20:18.060 --> 00:20:20.880
And from the dashboard, I can do things, like, I've got, like, my calendar in there.
00:20:20.880 --> 00:20:24.760
Like, if I went and looked at my agenda view today, it would have this podcast recording
00:20:24.760 --> 00:20:25.520
as a meeting on there.
00:20:25.920 --> 00:20:28.880
I can click on podcast recording, and then I've got a little button there that says,
00:20:29.360 --> 00:20:32.280
open your brief, and it'll just pop me open straight into Obsidian.
00:20:32.680 --> 00:20:39.400
Like, it takes me right into the vault, right to that thing, and gets me ready to go and
00:20:39.400 --> 00:20:40.700
where I need to be.
00:20:41.300 --> 00:20:44.800
So that's been super helpful and super contextual for me.
00:20:44.820 --> 00:20:49.540
Because when I come in the morning, I refresh my homepage for my WorkOS, right?
00:20:49.540 --> 00:20:51.520
I just hit the refresh button.
00:20:52.040 --> 00:20:54.860
And then all of a sudden, I'm presented with, here's your daily priorities.
00:20:55.260 --> 00:20:56.940
Here's your top three to-dos for the day.
00:20:57.160 --> 00:20:58.620
Here's your agenda for the day.
00:20:59.020 --> 00:21:02.540
Here's how these priorities relate to your agenda for the day.
00:21:02.760 --> 00:21:04.720
Here's all the things that I generated for you.
00:21:04.820 --> 00:21:06.040
Here's how many aren't left.
00:21:06.400 --> 00:21:06.860
Things like that.
00:21:07.260 --> 00:21:09.880
It's a very freeing thing.
00:21:10.160 --> 00:21:13.020
Like, in the past, I would have sat there and said, like, oh, crap, I've got a meeting.
00:21:13.180 --> 00:21:13.980
Let me go to that meeting.
00:21:14.400 --> 00:21:18.080
And you kind of come in all discombobulated, and you do the context switch, and you come
00:21:18.080 --> 00:21:18.660
out to the other side.
00:21:18.960 --> 00:21:20.600
Then you go, okay, I've got to go to my next meeting.
00:21:20.700 --> 00:21:24.500
I don't have time to go back and kind of write my notes up or capture them, anything
00:21:24.500 --> 00:21:25.060
like that.
00:21:25.440 --> 00:21:27.440
Now it's all just turnkey.
00:21:27.920 --> 00:21:32.820
It's all right there for me and ready to go.
00:21:33.280 --> 00:21:35.640
And also, very deterministic.
00:21:36.280 --> 00:21:42.060
So the last time that we chatted about this, I believe I had also mentioned kind of taking
00:21:42.060 --> 00:21:48.960
skills that maybe start out with a bunch of overarching shape in them, but then you want to make them
00:21:48.960 --> 00:21:50.540
more deterministic over time.
00:21:50.540 --> 00:21:54.680
So sometimes you have to stop relying on the model and the skill itself.
00:21:55.260 --> 00:22:00.220
And you need to kind of write scripts and have a script do the generation for you so that
00:22:00.220 --> 00:22:03.420
things are always formatted consistently, they come in the right way, all that.
00:22:04.080 --> 00:22:08.100
So for me, my meeting prep skill, it started as a very generic skill.
00:22:08.360 --> 00:22:13.380
Like, go find the transcript and the recording for this meeting on my calendar and capture it.
00:22:13.640 --> 00:22:16.480
And then I refined it over time, but it was still a bunch of markdown.
00:22:16.480 --> 00:22:21.760
I actually took it and worked and iterated on it back and forth a little bit, and I was able
00:22:21.760 --> 00:22:22.180
to get it.
00:22:22.360 --> 00:22:27.620
So now my meeting prep skill is really a script that calls WorkIQ.
00:22:28.120 --> 00:22:33.960
And it calls WorkIQ in a way that the output is very deterministic for me for every single
00:22:33.960 --> 00:22:34.740
meeting I go to.
00:22:34.940 --> 00:22:39.840
It's got the meeting attendees, it's got the transcript overview and the summary, it's got
00:22:39.840 --> 00:22:43.960
here's the key items from the chat, here's the documents that were shared, things like that.
00:22:44.520 --> 00:22:50.580
And what that allowed me to do by putting that skill, and now the skill is just a wrapper for
00:22:50.580 --> 00:22:53.800
calling that script, back to that dashboard thing, right?
00:22:53.840 --> 00:22:57.340
That dashboard's there, like I can't click a button on that dashboard and have it go run
00:22:57.340 --> 00:22:57.980
a skill.
00:22:58.240 --> 00:23:00.080
But I can have it go run a script, right?
00:23:00.120 --> 00:23:04.060
And kick that off very easily, like it just goes and runs a Python script in the background.
00:23:04.740 --> 00:23:09.880
So whether I'm running the skill through the harness, through Scout or through GitHub Copilot
00:23:09.880 --> 00:23:14.680
CLI, or I'm clicking the button, at the end of the day, the skill and me clicking the button
00:23:14.680 --> 00:23:20.480
are calling the same Python script, giving me the same deterministic output every single
00:23:20.480 --> 00:23:20.800
time.
00:23:20.880 --> 00:23:24.780
It's interesting to hear you talk because I've done some things very similar, but then
00:23:24.780 --> 00:23:26.220
I've also done some different things.
00:23:26.340 --> 00:23:28.620
Like I said, it's a very empowering thing, I think.
00:23:28.760 --> 00:23:32.460
So this is in no way, shape, or form a badge of honor.
00:23:32.800 --> 00:23:35.860
I'm actually kind of like scared when I say stuff like this out loud.
00:23:36.220 --> 00:23:39.920
I have on average 65 meetings a week on my calendar.
00:23:40.040 --> 00:23:40.800
That's nuts.
00:23:40.940 --> 00:23:41.860
So that's my average.
00:23:42.340 --> 00:23:49.020
Some weeks I have 80, some weeks I have 40, but I'm sitting right around 65 on average
00:23:49.020 --> 00:23:51.720
over the past three months since I started tracking this.
00:23:52.220 --> 00:23:55.380
I do not attend 65 meetings a week, to be very clear.
00:23:55.520 --> 00:23:59.100
Like 20 of those are probably just like learning or like FYI things.
00:23:59.600 --> 00:24:04.100
But having that skill out there, that prep and capture skill, like that still allows me to
00:24:04.100 --> 00:24:05.680
ingest the knowledge and have it out there.
00:24:05.760 --> 00:24:12.100
Like we had a training last week on like GitHub Copilot CLI and like sub-agent usage and things
00:24:12.100 --> 00:24:12.360
in it.
00:24:12.440 --> 00:24:13.320
I was like, oh, that's cool.
00:24:13.420 --> 00:24:14.580
Like I really want to go catch up with that.
00:24:15.000 --> 00:24:17.700
In the past, I would have completely forgotten about that, right?
00:24:17.720 --> 00:24:19.340
I would have had it written down on a piece of paper.
00:24:19.340 --> 00:24:23.020
I would have had like a to-do and to-do to go watch the recording for that.
00:24:23.360 --> 00:24:25.260
And I never would have gotten to it or come back around.
00:24:25.720 --> 00:24:30.080
But because I have this like prep and capture thing, it actually can sit there and contextualize
00:24:30.080 --> 00:24:34.540
for me and not only get me the information I need, so I can go get the summary and decide
00:24:34.540 --> 00:24:36.480
like, oh, was that actually worth my time or not?
00:24:36.960 --> 00:24:43.560
But then it's got all the canonical links back to the full recap and everything that I need
00:24:43.560 --> 00:24:45.360
out there and that's ready to go.
00:24:45.360 --> 00:24:50.220
So I'm probably like on some further end of the pendulum swinging.
00:24:50.440 --> 00:24:54.080
Like I haven't quite found equilibrium there where I'm probably still capturing too much
00:24:54.080 --> 00:24:56.480
stuff and I'm bringing in too much context.
00:24:56.680 --> 00:25:00.400
But I'm still at the point even like a couple months into this where I'd rather have more
00:25:00.400 --> 00:25:01.040
than less.
00:25:01.680 --> 00:25:04.960
And I hope I can get to like the less is more kind of thing along the way.
00:25:05.400 --> 00:25:10.360
But when it's so easy to do, because the loops are there and the harness is there, I'm kind
00:25:10.360 --> 00:25:11.600
of inclined just to keep it running.
00:25:12.280 --> 00:25:16.940
And then, like I said, it makes a bunch of this stuff like more powerful for me, right?
00:25:16.980 --> 00:25:23.280
Because if I'm going and like we're working on maybe like a new like CLI component or we're
00:25:23.280 --> 00:25:29.600
doing an MCP thing or some kind of like agentic thing and now I've got all these training on
00:25:29.600 --> 00:25:32.820
how agents work and how they do stuff, like that all applies, right?
00:25:32.880 --> 00:25:36.160
That all comes back into what I need to do.
00:25:36.360 --> 00:25:43.140
So contextually, that stuff could be part of my research for a SPAC or a PRD like that
00:25:43.140 --> 00:25:44.580
something later.
00:25:45.160 --> 00:25:49.000
And then I have these aspirations to go build some more loops.
00:25:49.500 --> 00:25:50.620
And I very much want to do this.
00:25:50.860 --> 00:25:58.740
So for me, for example, one of the things that I like to do is I like to go on Reddit and
00:25:58.740 --> 00:26:02.120
LinkedIn to go find feedback from users.
00:26:02.720 --> 00:26:04.000
So I've got a couple different places.
00:26:04.540 --> 00:26:07.080
Azure Storage really doesn't show up on Hacker News for some reason.
00:26:07.400 --> 00:26:08.880
They must not think we're a startup or something.
00:26:08.980 --> 00:26:09.220
I don't know.
00:26:09.620 --> 00:26:15.680
But we get like decent feedback on forums like Microsoft Q&A and Reddit and things like that.
00:26:15.900 --> 00:26:20.480
So I would love to go and build like a user research loop that can help me go like research
00:26:20.480 --> 00:26:24.520
my product on like a daily or weekly or monthly basis, things like that might be a little
00:26:24.520 --> 00:26:25.200
bit harder for me.
00:26:25.320 --> 00:26:26.360
I saw just today.
00:26:26.480 --> 00:26:27.740
So today is October 2nd.
00:26:28.240 --> 00:26:33.300
I think I saw earlier that Reddit is going to shut down all their external search access
00:26:33.300 --> 00:26:34.940
and all their RSS feed access.
00:26:35.460 --> 00:26:38.640
So that's going to make it a lot harder for agents to get at the data that's in Reddit.
00:26:38.920 --> 00:26:40.140
So maybe it's a good thing.
00:26:40.240 --> 00:26:44.360
I haven't built that loop and skill yet to go and do that.
00:26:44.880 --> 00:26:50.200
But the whole thing of starting with skills, automating those skills, and then figuring out
00:26:50.200 --> 00:26:54.940
how to compound them into these loops that just kind of keep running until you get to the
00:26:54.940 --> 00:26:57.080
outcome you want, super powerful.
00:26:57.080 --> 00:27:03.120
If you're managing file storage today, you've probably felt the pressure.
00:27:03.460 --> 00:27:05.300
Refresh cycles are harder to plan.
00:27:05.520 --> 00:27:10.420
Costs and complexity are increasing, especially as file workloads scale across regions.
00:27:10.900 --> 00:27:13.540
That's why Microsoft customers trust Nasuni.
00:27:13.820 --> 00:27:20.000
As an Azure certified software, Nasuni provides a global file system designed for enterprise scale
00:27:20.000 --> 00:27:25.700
environments, keeping file data secure, governed, and accessible across Azure environments without
00:27:25.700 --> 00:27:27.440
added infrastructure complexity.
00:27:27.440 --> 00:27:33.680
It's the foundation to keep your unstructured data stable, even as your infrastructure evolves.
00:27:33.940 --> 00:27:39.880
To see how much you could save while keeping your initiatives on track, visit nasuni.com
00:27:39.880 --> 00:27:50.040
I have a few that I run regularly.
00:27:50.040 --> 00:27:54.080
I have not gotten too far into the loops yet.
00:27:54.340 --> 00:27:57.800
I have a lot of stuff that I still just run manually.
00:27:57.800 --> 00:28:03.680
I have a few things that I run as routines or scheduled jobs.
00:28:04.100 --> 00:28:08.920
But it's been, yeah, it's a lot of the way I would look at these LLM wikis.
00:28:09.040 --> 00:28:14.000
And even again, it's probably taking it further than what the LLM wiki did, like with your dashboard
00:28:14.000 --> 00:28:14.600
and HTML.
00:28:15.020 --> 00:28:20.380
I actually built my own custom Obsidian plugin that I have as a dashboard.
00:28:20.820 --> 00:28:23.620
So my dashboard sits right in Obsidian rather than a web page.
00:28:23.820 --> 00:28:24.520
How did you do that?
00:28:24.760 --> 00:28:25.640
I'm intrigued.
00:28:25.920 --> 00:28:30.880
I told Claude, so I actually told Claude to go build me a custom Obsidian plugin.
00:28:31.300 --> 00:28:32.720
And Claude has built the...
00:28:32.720 --> 00:28:34.040
Okay, built it as a plugin.
00:28:34.220 --> 00:28:35.420
So it renders in the canvas.
00:28:35.600 --> 00:28:37.100
It built it as a plugin.
00:28:37.400 --> 00:28:41.800
So it renders as a canvas, like down the left side of Obsidian where you have your different
00:28:41.800 --> 00:28:42.220
icons.
00:28:42.700 --> 00:28:44.900
I now have an icon for my dashboard.
00:28:45.180 --> 00:28:47.300
And my dashboard has a bunch of cards on it.
00:28:47.560 --> 00:28:53.140
So this is where similar type of concept with mine, only I combined a bunch of stuff.
00:28:53.140 --> 00:28:57.700
I have multiple LLM wikis in the same vault because I have one for work.
00:28:57.940 --> 00:28:59.480
I have one for some personal stuff.
00:28:59.780 --> 00:29:01.940
I have a separate podcast one in my vault.
00:29:02.220 --> 00:29:10.280
Because sometimes for me, the overlap between a topic we talk about on a podcast and a blog
00:29:10.280 --> 00:29:15.200
post, maybe I want to write for my website and a client that I'm working with, there's so
00:29:15.200 --> 00:29:16.300
much overlap there.
00:29:16.300 --> 00:29:22.100
I want them to be separate to an extent, but I also want to be able to correlate certain
00:29:22.100 --> 00:29:22.860
things across them.
00:29:23.100 --> 00:29:26.200
But then my dashboard has cards that are color-coded and labeled.
00:29:26.640 --> 00:29:34.840
So I have intelligent cards for my top to-dos, maybe things that are on fire, CRM type information,
00:29:35.140 --> 00:29:38.160
potential jobs, where jobs are in the pipeline.
00:29:38.480 --> 00:29:45.720
I have some podcast stuff where I pull in podcast stats so I can see trends with the podcast
00:29:45.720 --> 00:29:49.100
downloads and episodes month to month.
00:29:49.420 --> 00:29:52.380
I have some personal stuff where I actually pull in stock market.
00:29:52.380 --> 00:29:58.620
I have a markdown file where I put stock tickers I want to track, and then I have a dashboard
00:29:58.620 --> 00:30:03.520
that every time I go to it, it'll go out and pull the latest stock prices and put it on
00:30:03.520 --> 00:30:04.060
the dashboard.
00:30:04.540 --> 00:30:10.380
But then I put little buttons on it that'll call a skill to go do analysis on that stock
00:30:10.380 --> 00:30:10.700
ticker.
00:30:10.800 --> 00:30:16.020
So I have my own analysis on stock tickers within my LLM wiki.
00:30:16.160 --> 00:30:18.760
And I might have to explore this because one of the things...
00:30:18.760 --> 00:30:19.560
It's interesting.
00:30:19.740 --> 00:30:23.700
I am definitely juggling between Scout, which sits out there.
00:30:23.780 --> 00:30:24.700
It's a desktop app, right?
00:30:24.720 --> 00:30:25.480
It has to be running.
00:30:25.980 --> 00:30:28.960
So I've got Scout sitting out there running and...
00:30:28.960 --> 00:30:29.820
Wow, that hurt.
00:30:30.100 --> 00:30:30.780
Sorry about that.
00:30:30.920 --> 00:30:33.660
It was quite a whine by that dog, by that puppy.
00:30:33.780 --> 00:30:35.180
I've got Scout out there running.
00:30:35.660 --> 00:30:37.760
I have this website that kind of runs.
00:30:37.840 --> 00:30:39.320
It's just a pin tab in Edge for me.
00:30:39.320 --> 00:30:41.580
It's always my first tab sitting at the top.
00:30:42.200 --> 00:30:44.940
And then I have Obsidian.
00:30:45.440 --> 00:30:49.440
So having just one less thing, like I don't have to go to the website and I can just live
00:30:49.440 --> 00:30:49.880
in Obsidian.
00:30:50.240 --> 00:30:53.160
Because I do this thing where I've been tempted to...
00:30:53.160 --> 00:30:54.320
Maybe I was thinking about it wrong.
00:30:54.380 --> 00:30:59.400
I was actually tempted to see if I could find a way to embed a Markdown viewer in an Obsidian
00:30:59.400 --> 00:31:00.440
view in my website.
00:31:00.700 --> 00:31:02.200
But I should just totally go the other way.
00:31:02.720 --> 00:31:04.120
The Canvas thing is very interesting.
00:31:04.120 --> 00:31:06.060
So I'm glad you brought that up.
00:31:06.140 --> 00:31:09.360
I've seen a bunch of people doing similar things recently.
00:31:10.100 --> 00:31:16.340
So if you're in VS Code and you're using, say, like GitHub Copilot chat in VS Code.
00:31:16.960 --> 00:31:19.740
So VS Code has plugins, all those kinds of things.
00:31:19.840 --> 00:31:24.580
You can absolutely use your LLM to help you develop a plugin that draws on the canvas in
00:31:24.580 --> 00:31:25.160
VS Code.
00:31:25.520 --> 00:31:27.540
The other one is GitHub Copilot.
00:31:27.540 --> 00:31:29.540
So GitHub Copilot has a CLI.
00:31:29.960 --> 00:31:32.820
And there's a new GitHub Copilot app as well.
00:31:33.320 --> 00:31:37.020
And that app renders canvases in HTML as well.
00:31:37.540 --> 00:31:43.140
So I've seen some folks in my organization that are creating custom canvas apps for GitHub
00:31:43.140 --> 00:31:45.680
Copilot, the desktop app, because that's where they're living.
00:31:46.180 --> 00:31:49.000
And that's where they're doing their work out of kind of thing.
00:31:49.140 --> 00:31:54.640
So that's another super powerful concept is like kind of coaxing an LLM and getting it
00:31:54.640 --> 00:32:00.020
to build the view that you want and the experience you want in the place that you work.
00:32:00.280 --> 00:32:03.080
So going back to your skills, I have a lot of similar skills.
00:32:03.080 --> 00:32:08.420
Like I have a daily review skill that will go back and look at all my meeting notes from
00:32:08.420 --> 00:32:09.020
the day.
00:32:09.200 --> 00:32:10.680
It'll look at my calendar.
00:32:11.260 --> 00:32:15.240
Mine actually goes into, I don't know if you've tried this, it'll go into my team's
00:32:15.240 --> 00:32:15.840
conversations.
00:32:16.480 --> 00:32:22.560
And I've told it to look for certain, there's certain teams that I'm in that are super chatty
00:32:22.560 --> 00:32:25.760
where only like five or 10% of the stuff is relevant to me.
00:32:26.000 --> 00:32:32.480
And I'm like, go look through all of these chats and teams and give me, tell me where there's
00:32:32.480 --> 00:32:36.620
a chat related to a certain topic because that's the only thing I care about.
00:32:36.920 --> 00:32:41.540
So part of my review will actually analyze all my team's chats for me and tell me if there's
00:32:41.540 --> 00:32:45.120
anything I should go back and look at my team's chats or if I can just ignore all of them.
00:32:45.120 --> 00:32:46.040
I run a similar one.
00:32:46.380 --> 00:32:49.400
That's another automation or loop that I have out there running.
00:32:50.200 --> 00:32:53.500
So it looks across my inbox and my team's chats.
00:32:53.620 --> 00:32:54.080
And your team's.
00:32:54.080 --> 00:33:00.480
And it looks for emails from my manager, my skip, or people on my team.
00:33:00.540 --> 00:33:00.760
Okay.
00:33:00.900 --> 00:33:01.700
Who report to me.
00:33:02.560 --> 00:33:10.600
And initially that was a, let's get out there and kind of grab everything, right?
00:33:10.640 --> 00:33:13.700
Like the signal to noise ratio was, it was really bad.
00:33:13.700 --> 00:33:18.320
It was pretty horrible because like anytime somebody chatted with me on my team, like they
00:33:18.320 --> 00:33:20.960
could ping me on the side and say, how was the weather today?
00:33:21.200 --> 00:33:26.580
And I would all of a sudden get this alert that popped up and said, Joe is, Joe is, I
00:33:26.580 --> 00:33:30.280
don't have Joe on my team, but Joe is asking you how the weather is today.
00:33:30.380 --> 00:33:31.180
Do you want to respond?
00:33:31.520 --> 00:33:33.180
Oh no, that's, that's not important, right?
00:33:33.660 --> 00:33:36.100
So I definitely had to go and refine that thing in a similar way.
00:33:36.400 --> 00:33:39.380
I have a little bit of contextual filter, things like that.
00:33:39.380 --> 00:33:45.600
So it only alerts me now when folks are kind of pinging me with urgency or high priority
00:33:45.600 --> 00:33:51.100
items, particularly things that align to work streams that were kind of
00:33:51.100 --> 00:33:53.600
integrated on or that I need to know about what's happening.
00:33:54.100 --> 00:33:56.980
I've actually been thinking, and one of the things I would love to do is take it to the
00:33:56.980 --> 00:34:00.300
next level and just have it be able to respond back on teams.
00:34:00.920 --> 00:34:08.120
So like if somebody says, oh, hey, can you give me the latest update on feature ABC?
00:34:08.880 --> 00:34:13.380
If I already sent out the status update for feature ABC this week, I just wanted to send
00:34:13.380 --> 00:34:19.360
them the doc or send them a link like, or send them like the subject of that email that went
00:34:19.360 --> 00:34:20.300
out or whatever that was.
00:34:20.860 --> 00:34:26.340
And that feels like low hanging fruit and something that I could go out there and automate fullness
00:34:26.340 --> 00:34:26.900
of time.
00:34:27.260 --> 00:34:29.000
I just haven't been sure if it's worth it yet.
00:34:29.320 --> 00:34:33.520
Like putting the effort in to have an additional contextual filter, definitely worth it.
00:34:33.560 --> 00:34:37.300
Like I said, like I was getting all sorts of like stuff, like very much along the lines
00:34:37.300 --> 00:34:40.820
of like, so-and-so is telling you the weather's nice today.
00:34:41.020 --> 00:34:41.940
Like very important.
00:34:42.100 --> 00:34:42.740
Please respond.
00:34:43.100 --> 00:34:44.200
Like, no, I don't need to do that.
00:34:44.200 --> 00:34:45.020
I don't find mine too.
00:34:45.120 --> 00:34:46.360
It's just certain topics.
00:34:46.540 --> 00:34:51.220
But then in my dashboard, so I can run all of my skills from my dashboard.
00:34:51.580 --> 00:34:58.280
I can just click a button in the dashboard and it'll go open a bash prompt, a Claude session
00:34:58.280 --> 00:35:03.100
in the background and go run the skill in the background and I won't ever see it.
00:35:03.180 --> 00:35:07.920
So any skills that I don't have to interact with, then I have other skills where I know
00:35:07.920 --> 00:35:13.420
like I may go back and forth with them on or have more of a conversation with Claude.
00:35:13.640 --> 00:35:16.420
Those, the button actually copies a prompt to my clipboard.
00:35:16.520 --> 00:35:21.000
So I just open up a Claude session and paste the prompt right into a Claude session.
00:35:21.100 --> 00:35:22.720
Interesting that you chose to go down that path.
00:35:22.820 --> 00:35:24.140
Everything out of my dashboard.
00:35:24.300 --> 00:35:30.040
Go open the harness, go open Claude and run the skill versus me where I was like, I'm going
00:35:30.040 --> 00:35:34.020
to go, I'm going to go have it write the script and have it be super deterministic so I can
00:35:34.020 --> 00:35:38.440
run that script from a skill or from a button or from here.
00:35:38.540 --> 00:35:40.860
Some of my skills do still run scripts.
00:35:41.260 --> 00:35:42.300
It's the same.
00:35:42.300 --> 00:35:45.960
I still have a lot of scripts where those skills, it's still deterministic.
00:35:46.080 --> 00:35:49.960
It's a button, open a Claude session, run a script, run a skill.
00:35:49.960 --> 00:35:53.580
So one of the things maybe I was going at and something for folks to think about is, and
00:35:53.580 --> 00:35:58.780
one of the reasons I kind of go down this path quite a bit, is optimizing for token usage
00:35:58.780 --> 00:36:00.260
and context.
00:36:01.100 --> 00:36:07.040
So if you're using Claude as a harness, Claude has a fairly large system context already coming
00:36:07.040 --> 00:36:08.860
with it as part of the harness.
00:36:08.860 --> 00:36:13.280
And every time you're invoking a skill, you have to load that context no matter what.
00:36:13.280 --> 00:36:14.900
Like that's there.
00:36:14.900 --> 00:36:19.940
And so sure, you should have the advantages of cash tokens and things like that.
00:36:19.940 --> 00:36:24.500
If Anthropik's doing their job right, if they're not doing their job right on a given day, or
00:36:24.500 --> 00:36:29.000
stuff does go sideways, you're really in this place where you're burning tokens.
00:36:29.000 --> 00:36:33.240
And this stuff does cost money, right?
00:36:33.380 --> 00:36:37.500
Like, I'd be curious if you're comfortable saying like, do you have a sense for how much
00:36:37.500 --> 00:36:40.540
yours costs to run over the course of a month?
00:36:40.680 --> 00:36:41.040
If not?
00:36:41.120 --> 00:36:42.620
I do other things in it.
00:36:42.740 --> 00:36:47.580
And I mean, I will say I am paying for Claude team now.
00:36:47.840 --> 00:36:51.040
So I'm paying for the team edition of Claude.
00:36:51.160 --> 00:36:52.380
What's the team edition run?
00:36:52.560 --> 00:36:58.240
Which it's somewhere between, so they have like the Claude max, where it's the 5x or the
00:36:58.240 --> 00:36:58.840
20x.
00:36:58.960 --> 00:37:02.440
I'm actually sitting at like 6.5x.
00:37:02.440 --> 00:37:05.200
They don't have a 5x or 20x.
00:37:05.500 --> 00:37:08.680
So I would say I'm equivalent to the 5x, but it's a little bit more.
00:37:08.780 --> 00:37:11.120
I'm paying $125 a month for Claude.
00:37:11.200 --> 00:37:16.100
So my token consumption input and output is 100 to 120.
00:37:16.560 --> 00:37:20.720
I don't have the hard average in front of me, but I would say I probably average.
00:37:20.840 --> 00:37:26.420
I probably average like $110, $115 a month, something like that, to go through and run
00:37:26.420 --> 00:37:27.780
everything that I need to do.
00:37:28.340 --> 00:37:33.100
But that includes like the things like, hey, going and like bootstrapping the dashboard and
00:37:33.100 --> 00:37:33.660
things like that.
00:37:33.840 --> 00:37:37.980
I would expect like continued optimization as things come down.
00:37:38.120 --> 00:37:39.540
I was trying to find this the other day.
00:37:39.780 --> 00:37:44.060
Over the last 30 days, I can see things like product stickiness.
00:37:44.260 --> 00:37:45.000
This is connectors.
00:37:45.300 --> 00:37:46.680
I can see the skills.
00:37:47.620 --> 00:37:52.720
Most of these run in Claude code, usage, top skills, tool usage.
00:37:53.000 --> 00:37:59.580
I was just trying to look super quick if I could see how many gives me value.
00:38:00.280 --> 00:38:01.540
I haven't found.
00:38:02.100 --> 00:38:06.820
I'd have to dig through here like how many tokens I use.
00:38:07.040 --> 00:38:07.800
It's not free.
00:38:08.480 --> 00:38:09.500
No, it's not free.
00:38:09.560 --> 00:38:13.740
But I can fit it all in there with room to spare because I run these skills, but I also
00:38:13.740 --> 00:38:18.280
use it all the time for writing PowerShell scripts or analyzing data, that type of stuff.
00:38:18.620 --> 00:38:25.520
Because I do, similar to you, I have a lot of these skills that have turned into Python
00:38:25.520 --> 00:38:31.800
scripts in the background as well, where the skill calls a script to do a lot of the work
00:38:31.800 --> 00:38:32.160
for it.
00:38:32.280 --> 00:38:34.060
Python seems to be where it all lands.
00:38:34.280 --> 00:38:38.520
But a lot of what I do is very similar type of stuff that I built to you, except that
00:38:38.520 --> 00:38:41.180
I have multiple LLM wikis in that vault.
00:38:41.500 --> 00:38:43.760
And then I did the dashboard in Obsidian.
00:38:44.940 --> 00:38:49.500
It's been, at least from my end, super empowering because I get to work the way I want to work.
00:38:49.760 --> 00:38:55.180
I'm always frustrated at the friction of, here's Microsoft to do, and it does to do the way
00:38:55.180 --> 00:39:00.020
Microsoft wants to do them for that product or things like that.
00:39:00.360 --> 00:39:06.620
So I would love it if I could say, I could take this thing and package it up and give it
00:39:06.620 --> 00:39:08.480
to you and it would work for the way you work.
00:39:08.720 --> 00:39:11.600
I don't think that's the case, nor do I think it's the point.
00:39:11.820 --> 00:39:18.600
The whole point is to have the way you work and the set of skills and loops and sub-agents
00:39:18.600 --> 00:39:25.540
or authentic flows and connectors and everything else that are out there to wire everything
00:39:25.540 --> 00:39:28.540
up and get it going and have a good set of outcomes for it.
00:39:28.680 --> 00:39:32.900
I really do feel like it helps accelerate the work that I do.
00:39:33.440 --> 00:39:35.880
I think it keeps me better grounded.
00:39:36.580 --> 00:39:39.980
I still have to go through a lot of context switching and I find it super helpful to have
00:39:39.980 --> 00:39:43.260
a lot of the grounded notes and things just right there at my fingertips.
00:39:43.260 --> 00:39:49.420
Go ask a question so I can search in Obsidian or when I do need to use an LLM to iterate
00:39:49.420 --> 00:39:51.020
over it, great, here you go.
00:39:51.120 --> 00:39:58.420
I can hop in that directory and invoke my harness of choice and I'm off to the races.
00:39:58.680 --> 00:40:02.840
I did things a little differently too for my tasks because I have contractors working for
00:40:02.840 --> 00:40:03.040
me.
00:40:03.360 --> 00:40:08.640
So I want to make sure tasks are updated in Asana and some of that.
00:40:08.640 --> 00:40:14.940
So my dashboard, instead of managing tasks here, all of my skills and prompts and everything
00:40:14.940 --> 00:40:17.740
ensure that Asana stays the source of truth.
00:40:18.080 --> 00:40:23.920
So if I see something on my list of my top to do or something that I need to get done, it
00:40:23.920 --> 00:40:29.340
actually links to Asana so I can click it, open it in Asana and I check it off there and
00:40:29.340 --> 00:40:31.420
then the dashboard will update.
00:40:31.980 --> 00:40:33.780
And same thing for on fire.
00:40:33.900 --> 00:40:36.900
I have two right now where there's an email and a Teams conversation.
00:40:37.080 --> 00:40:42.660
I have created my dashboard so I can click on those items that I've pulled in and it takes
00:40:42.660 --> 00:40:46.460
me right to the email or it takes me right to the Teams conversation.
00:40:46.840 --> 00:40:53.200
I have one for blog posts that takes me right to the draft blog posts or to the client record
00:40:53.200 --> 00:40:54.160
in my CRM.
00:40:54.280 --> 00:41:00.420
So my dashboard, the way I've thought about mine, I've turned my dashboard and my LLM into
00:41:00.420 --> 00:41:07.060
an integration point so I can pull all the information from all the systems I want and built a central,
00:41:07.080 --> 00:41:12.760
dashboard command center that I can work from but then link out to all the different
00:41:12.760 --> 00:41:15.720
things that I need to respond to that pop up on it.
00:41:15.800 --> 00:41:17.540
I'm on a very similar path.
00:41:17.740 --> 00:41:22.220
I have this kind of backlog of things that I want to go kind of integrate next.
00:41:22.460 --> 00:41:28.760
So one is, I mentioned we use Azure DevOps, ADO, for kind of tracking of epics and features
00:41:28.760 --> 00:41:30.420
and PBIs and things like that.
00:41:30.420 --> 00:41:36.340
I'd love to have that integrated view right there for, hey, I'm working on project XYZ,
00:41:36.720 --> 00:41:42.340
click on project XYZ, here's the ADO items for it, and here's their current state and status,
00:41:42.980 --> 00:41:44.460
things like that.
00:41:44.740 --> 00:41:50.260
The other one that I want to build is, at some point, I struggle with this stuff, I don't
00:41:50.260 --> 00:41:51.460
know when I'm reinventing the wheel.
00:41:52.260 --> 00:41:56.980
This one feels a little bit maybe like reinventing SharePoint, but I've got these projects that
00:41:56.980 --> 00:41:57.300
exist.
00:41:57.680 --> 00:42:02.720
So when you do the Carpathia-style LLM wikis, you end up with artifacts and entities, you end
00:42:02.720 --> 00:42:07.720
up with these discrete templates for types of objects in your LLM wiki.
00:42:08.180 --> 00:42:13.320
So I've got projects and work streams and events and things like that.
00:42:13.920 --> 00:42:18.880
So for my work streams and my projects, there's these kind of canonical briefs that sit there
00:42:18.880 --> 00:42:21.240
that have all the detail about them.
00:42:21.760 --> 00:42:24.800
And then they have all the wiki links to sublinks and things like that.
00:42:25.240 --> 00:42:30.440
But sometimes the sublinks they carry are links to documents, so like Word docs, Excels,
00:42:30.620 --> 00:42:31.220
PowerPoints.
00:42:31.220 --> 00:42:39.680
So I kind of want to view where I can say, hey, for Workstream ABC, or Workstream A, click
00:42:39.680 --> 00:42:42.480
on Workstream A, here's all the important documents for it.
00:42:42.580 --> 00:42:46.340
Like, here's all the contextual things and just like the little snippet and summary of
00:42:46.340 --> 00:42:46.820
it that's there.
00:42:47.100 --> 00:42:50.680
And then like the document, to be very clear, like the artifacts, the canonical artifact,
00:42:50.900 --> 00:42:51.920
not stored in my wiki.
00:42:52.140 --> 00:42:56.220
Like that stuff's still IRM, it's got sensitivity labels, all that stuff.
00:42:56.440 --> 00:42:58.220
But the link to it is in my wiki.
00:42:58.220 --> 00:43:02.840
So now it's like one-stop shopping for me, where in the past, I would have been going
00:43:02.840 --> 00:43:08.360
to SharePoint and just going to kind of like the generic intranet search or even like Copilot
00:43:08.360 --> 00:43:08.860
search.
00:43:09.020 --> 00:43:12.740
I'm finding it to be kind of annoying because it's an LLM-based search.
00:43:12.920 --> 00:43:16.800
So you kind of start talking to it like an LLM and it goes sideways pretty quick.
00:43:17.520 --> 00:43:19.040
So I've got all this stuff sitting.
00:43:19.340 --> 00:43:20.480
Like the context is there.
00:43:20.640 --> 00:43:25.320
It's just about, like you said, integrating the views and kind of making it one-stop shopping
00:43:25.320 --> 00:43:27.100
across all these various systems.
00:43:27.100 --> 00:43:32.940
Yeah, my project for my dashboard has a backlog markdown file in it where I just, when I come
00:43:32.940 --> 00:43:35.160
up with an idea, I throw an idea in there.
00:43:35.340 --> 00:43:38.760
So when I have some time or I'm sitting on the couch in the evening and I want to build
00:43:38.760 --> 00:43:42.880
a new feature on my dashboard, I have my whole backlog and I can pick one off the backlog
00:43:42.880 --> 00:43:43.720
to start building.
00:43:43.840 --> 00:43:44.640
Anything we miss.
00:43:44.860 --> 00:43:47.900
I'm sure this is one we're going to come back to because I feel like, to your point,
00:43:47.980 --> 00:43:52.940
you're always evolving it, changing it, modifying it as everything changes.
00:43:53.240 --> 00:43:54.800
And as you come up with ideas.
00:43:54.800 --> 00:43:58.400
The builder's itch comes in along the way.
00:43:58.800 --> 00:44:03.980
And then the other thing I like about it is, so you not only get to go build these things,
00:44:04.280 --> 00:44:07.220
but like I said, you get to adapt them to the way you work.
00:44:07.700 --> 00:44:11.500
And you also get to adapt them to your taste.
00:44:12.040 --> 00:44:19.000
Like I said, I don't have the greatest reaction sometimes when I have to go into a Microsoft
00:44:19.000 --> 00:44:24.420
to-do or a planner, but it doesn't rub me nearly as much the wrong way going into this
00:44:24.420 --> 00:44:29.060
thing that I built for the way I guided an LM to build it for me, because there's no way
00:44:29.060 --> 00:44:32.660
I was writing all that stuff myself to go and put it together.
00:44:32.660 --> 00:44:36.020
So it actually, it's like, it's a little bit more calming, right?
00:44:36.100 --> 00:44:40.300
Like while it's bringing in probably, like I said, more stuff.
00:44:40.520 --> 00:44:46.480
And if anything, like it's allowing me to over-function, but it's at least allowing me to over-function
00:44:46.480 --> 00:44:51.880
in like my way of over-functioning without like my head constantly on a swivel.
00:44:52.080 --> 00:44:56.900
That's the biggest advantage I've found of this and frustrations I've had other times
00:44:56.900 --> 00:45:01.820
that I'm like, how do I integrate Asana with Teams, with Outlook, in a way that works for
00:45:01.820 --> 00:45:02.820
how I want to work?
00:45:03.180 --> 00:45:09.940
This has allowed me to pull that information and put some integrations in place that A,
00:45:09.980 --> 00:45:12.380
I can never develop or build on my own.
00:45:12.780 --> 00:45:18.060
And B, I can use products, but then tie them all together in a way that makes sense for
00:45:18.060 --> 00:45:22.000
the way I want to see information, work with information, manage my day.
00:45:22.080 --> 00:45:22.700
Et cetera.
00:45:23.100 --> 00:45:27.120
So that's the part I've enjoyed about this, to your point.
00:45:27.380 --> 00:45:29.260
It's, I can tweak it to me.
00:45:29.600 --> 00:45:31.260
So with that, just wrap it up.
00:45:31.340 --> 00:45:32.840
We went almost an hour today, Scott.
00:45:33.040 --> 00:45:34.020
Sometimes we go a little bit longer.
00:45:34.600 --> 00:45:37.140
So I'm always interested in feedback from other folks.
00:45:37.560 --> 00:45:41.900
We've mentioned a couple of times and it's good to see people coming and commenting on
00:45:41.900 --> 00:45:43.140
Reddit or LinkedIn.
00:45:43.720 --> 00:45:46.680
I would love to hear what other folks are doing there.
00:45:46.680 --> 00:45:52.780
Like, what's your superpower, your unlock for LLM wikis?
00:45:53.020 --> 00:45:54.340
Like, are you using them?
00:45:54.760 --> 00:45:59.780
I think you and I are maybe a little bit more on like one end of the spectrum where we're
00:45:59.780 --> 00:46:04.420
like in CLI harnesses or a little bit more advanced thing.
00:46:04.560 --> 00:46:11.020
Like, I've seen folks on definitely like YouTube using very much like Claude Cowork style kinds
00:46:11.020 --> 00:46:14.280
of things to go and execute these same workflows.
00:46:15.320 --> 00:46:20.720
So I'd love to understand where others are experimenting, what they're doing, where they're finding the
00:46:20.720 --> 00:46:25.540
unlock, and maybe what else we're missing that could be like the take it up another notch.
00:46:25.960 --> 00:46:27.380
We'll do the final tap thing.
00:46:27.540 --> 00:46:28.460
Like, does this go to 11?
00:46:28.680 --> 00:46:29.620
Yes, yes, it does.
00:46:29.760 --> 00:46:30.000
Yes.
00:46:30.160 --> 00:46:30.740
Let us know.
00:46:31.000 --> 00:46:31.700
We'd love to hear it.
00:46:31.940 --> 00:46:32.220
All right.
00:46:32.420 --> 00:46:33.060
Well, that's Scott.
00:46:33.260 --> 00:46:34.160
It's time for the weekend.
00:46:34.520 --> 00:46:36.000
Almost time for the weekend.
00:46:36.420 --> 00:46:37.060
But yeah.
00:46:37.620 --> 00:46:39.040
Well, as always, thanks, Ben.
00:46:39.300 --> 00:46:39.740
Not a problem.
00:46:40.000 --> 00:46:40.400
Thank you.
00:46:40.540 --> 00:46:41.380
And we'll talk to you again soon.
00:46:43.480 --> 00:46:47.300
If you enjoyed the podcast, go leave us a five star rating in iTunes.
00:46:47.300 --> 00:46:52.420
It helps to get the word out so more IT pros can learn about Office 365 and Azure.
00:46:53.340 --> 00:46:57.460
If you have any questions you want us to address on the show or feedback about the show,
00:46:57.900 --> 00:47:01.400
feel free to reach out via our website, Twitter, or Facebook.
00:47:01.860 --> 00:47:03.920
Thanks again for listening and have a great day.
