Leela Launch: Preeti Krishnan on Building Compliance That Runs Itself

WEBVTT
00:00:00.080 --> 00:00:01.600
Welcome to First Builders, everyone.
00:00:01.760 --> 00:00:10.400
We're doing a special off-season episode today with one of our newest portfolio companies that's announcing their launch and oversubscribed pre-seed raise, Leela.
00:00:10.960 --> 00:00:13.919
We're super excited to have been an early commitment into their round.
00:00:14.160 --> 00:00:16.640
And if you're tuning in for the first time, we're Council Capital.
00:00:16.719 --> 00:00:21.920
We invest in early stage companies in their first round of capital, and we are obsessed with first builders turned founders.
00:00:22.079 --> 00:00:26.879
These are people who joined previous companies prior to success and helped build through key inflection points.
00:00:27.039 --> 00:00:29.280
They're second time founders hidden in plain sight.
00:00:29.359 --> 00:00:32.960
So we're so excited to welcome Preeti Krishnan from Leela onto the show today.
00:00:33.119 --> 00:00:34.159
Welcome, Preet Preeti.
00:00:34.479 --> 00:00:35.679
Thank you both for having me.
00:00:35.759 --> 00:00:38.079
I'm so excited to be here, Amber and Rachel.
00:00:38.159 --> 00:00:40.640
And I'm so, so excited to be a part of the portfolio.
00:00:40.880 --> 00:00:41.759
Yeah, absolutely.
00:00:41.920 --> 00:00:42.960
We're thrilled as well.
00:00:43.119 --> 00:00:48.399
And we would love for you first, like one of the most exciting things, obviously, with our thesis was your background going into this.
00:00:48.479 --> 00:00:54.560
So I'd love for you to share, you know, what you were doing prior to launching Leela and how you discovered the problem.
00:00:54.880 --> 00:00:56.000
Yeah, absolutely.
00:00:56.240 --> 00:00:57.439
Thank you for mentioning that.
00:00:57.520 --> 00:01:08.239
I I think one of the things that was so exciting for me about partnering with you all was you really, I think, reflected back at me some of the reasons that I built this business in the first place.
00:01:08.480 --> 00:01:12.239
And so much of it comes from what I saw in the in the years before this.
00:01:12.400 --> 00:01:14.480
So I started my career as an attorney.
00:01:14.560 --> 00:01:16.959
I worked at a firm called Davis Polk for many years.
00:01:17.120 --> 00:01:18.400
I was in the funds group.
00:01:18.640 --> 00:01:28.480
And that was sort of the first time that I saw the complex web of opportunities, but also challenges that exist in regulated businesses and their compliance obligations.
00:01:28.719 --> 00:01:34.239
So at Davis Polk, what we would do is set up these funds for success, right?
00:01:34.319 --> 00:01:37.519
Um, we told them what obligations or rules they needed to follow.
00:01:37.680 --> 00:01:41.840
We made sure that they were structurally sound, and then we sent them off to the races.
00:01:42.079 --> 00:01:48.799
But what I saw over and over again was that's when a lot of the compliance work actually began for these firms, right?
00:01:49.040 --> 00:02:00.799
You know what rules you have to follow, but now you got to figure out how to follow them, how they fit your specific operations, your specific company, what that translates into, how you develop regimes around that.
00:02:01.040 --> 00:02:08.080
After spending some time there, I had the entrepreneurial sort slash startup uh bug hit me.
00:02:08.240 --> 00:02:12.879
And so I spent the majority of my career building startups um in the fintech space.
00:02:13.039 --> 00:02:18.319
So I first spent uh a lot of time, almost five years, at a company called Daily Pay.
00:02:18.400 --> 00:02:31.759
I started as chief of staff and then I moved into senior operating roles ultimately as COO of the company, as the company scaled from 20 employees and 2 million in ARR to a thousand employees and 250 million in ARR.
00:02:31.919 --> 00:02:45.840
And, you know, as as well as being a wild ride, it was also sort of the first time that I saw all of the operational complexity of compliance, kind of on the other side of where I had, you know, sat previously.
00:02:46.080 --> 00:02:53.919
And this time, the thing that I really picked up on was man, it is so hard for compliance operators to actually do their jobs.
00:02:54.080 --> 00:03:07.439
They are really obligated or, you know, imbued with the responsibility of seeing what's going on across the business, ensuring they're on top of it, making sure that everyone follows the rules without slowing the business down.
00:03:07.919 --> 00:03:11.039
And it always felt like an impossibility to me.
00:03:11.199 --> 00:03:18.960
Again, as an operator, I didn't want to spend my time, right, like dragging revenue down, slowing it down with compliance.
00:03:19.199 --> 00:03:23.039
And yet I knew every day, right, like what other option did they have?
00:03:23.199 --> 00:03:25.520
It was this like kind of crazy conundrum.
00:03:25.759 --> 00:03:38.000
After that, I spent some time at a lending business called Caffeine in the again alternative lending space, where compliance obligations were similarly so important to run our business.
00:03:38.159 --> 00:03:46.159
And yet the resources and the way in which it worked always meant that the compliance operator was wondering, what am I missing?
00:03:46.319 --> 00:03:59.120
And so I saw an opportunity, especially as technology started evolving, not just to give compliance operators resources and data to do their work, but actually help do the work for them ourselves.
00:03:59.360 --> 00:04:01.919
And so that was sort of where Leela blossomed from.
00:04:02.080 --> 00:04:10.560
I was frankly sitting on both sides of this problem, both the origination of compliance work as well as observing how hard it was to do it.
00:04:10.719 --> 00:04:18.399
And that sort of, you know, created the right set of questions for me to start putting the idea behind the company together.
00:04:18.639 --> 00:04:18.879
Yeah.
00:04:18.959 --> 00:04:24.399
I mean, I remember seeing this a lot at my time at Komodo Health, being in like the healthcare data industry as well.
00:04:24.480 --> 00:04:26.079
So very highly regulated.
00:04:26.399 --> 00:04:29.600
Um, but we wanted to move really fast, grow really fast.
00:04:30.160 --> 00:04:35.600
And oftentimes, you know, it was like very much like this push-pull kind of like with compliance.
00:04:35.680 --> 00:04:41.279
Like we knew we had to make sure that we were compliant and it was critically important given the data that we were working with.
00:04:41.439 --> 00:04:46.319
But yeah, we didn't want to be, you know, also slowing down revenue opportunities either.
00:04:46.480 --> 00:04:48.800
So it was definitely an interesting, I think, dynamic.
00:04:48.959 --> 00:04:53.199
And I'm sure all of these other regulated industries have like a similar dynamic as well.
00:04:53.439 --> 00:04:53.920
Absolutely.
00:04:54.079 --> 00:05:06.879
And I, you know, I think I'll just say on top of that, right, Rachel, like I remember at Davis Polk, the thing that we were told more than anything was legal and compliance exists to help regulated businesses grow.
00:05:07.120 --> 00:05:08.879
Like, how crazy is that, right?
00:05:08.959 --> 00:05:12.319
To actually think that the function exists to enable.
00:05:12.560 --> 00:05:14.879
And yet, so many of our experiences with it.
00:05:15.360 --> 00:05:16.800
It feels like the opposite negative.
00:05:17.040 --> 00:05:20.639
It's like, oh, it's slowing down, it's slowing down, but it can really open up opportunities.
00:05:20.879 --> 00:05:21.120
Yes.
00:05:21.199 --> 00:05:24.319
And the operators themselves, right, on the compliance side, know that.
00:05:24.480 --> 00:05:26.480
Like they don't want to slow the business down, right?
00:05:26.560 --> 00:05:31.519
Like their goal is to help you do more, use it as a competitive advantage, so on and so forth.
00:05:31.759 --> 00:05:34.240
But the manner by which it operates today, right?
00:05:34.319 --> 00:05:51.279
Like, not only is it heavily manual in terms of stitching together data, being able to drive information, all of that, but also just the amount of stuff, right, that comes through is nearly impossible without having to constantly go back in the business on a regular basis, right?
00:05:51.360 --> 00:05:55.759
So it's sort of this like series of unfortunate events that has come together.
00:05:56.079 --> 00:06:01.040
None without, you know, with intentionality, like everyone is trying to do their best, right?
00:06:01.199 --> 00:06:07.920
But we think that technology has finally gotten to a point where we can really make a step change in the way that the function exists.
00:06:08.079 --> 00:06:08.480
Yeah.
00:06:08.720 --> 00:06:12.319
And what have what have other tools not had that that Leela's doing?
00:06:12.399 --> 00:06:17.199
I would love to hear a little bit more about Leela and and what you're doing about to help with some of those problems.
00:06:17.680 --> 00:06:18.879
Thank you for asking, Amber.
00:06:18.959 --> 00:06:23.199
So if you think about the idea of what compliance is in the first place, right?
00:06:23.360 --> 00:06:32.879
And we focus on regulated compliance specifically, but we've seen these patterns and trends across lots of different compliance surfaces, but particularly in regulated compliance.
00:06:33.040 --> 00:06:42.240
Your lawyers tell you, you know, what laws apply, generally what you got to do about it, and then they, you know, send you off for the races, like I kind of described in in my old life.
00:06:42.399 --> 00:06:45.519
Then you have to figure out how all of it applies to you, right?
00:06:45.759 --> 00:06:57.040
The first step for any company company is typically to translate all of those regulatory, you know, provisions into policies that apply to your company specifically, right?
00:06:57.199 --> 00:06:58.480
What do we have to do?
00:06:58.720 --> 00:07:11.920
Many companies in regulated spaces have upwards of 40 policies, including one of our first customers, but you then have to translate all of those policies into their um component obligations, right?
00:07:12.079 --> 00:07:14.399
Things that you have to do as a company.
00:07:14.560 --> 00:07:16.079
But then you got to do the work.
00:07:16.240 --> 00:07:23.120
Those obligations require you to actually perform certain tasks that enable you to say, I was compliant, right?
00:07:23.279 --> 00:07:24.319
I followed the rules.
00:07:24.800 --> 00:07:28.000
The fourth piece is actually producing that evidence.
00:07:28.240 --> 00:07:35.199
I had a policy, I actually put those obligations into my business, I did the work, and now I'm showing you that I did it.
00:07:35.439 --> 00:07:41.360
This can be to a counterparty, to a regulator, to anyone who you owe some compliance obligation to.
00:07:41.600 --> 00:07:48.959
Now we've found that in the past, many compliance tools can solve components of various pieces of that stack.
00:07:49.199 --> 00:07:52.240
Maybe you have an AML obligation, right?
00:07:52.720 --> 00:08:09.839
If you do, there are tons and tons of wonderful tools that will enable you to scan all of your transaction data and see what obligations may or may not be fraudulent or follow or break the rules that you have or that are macro rules that you think about, so on and so forth.
00:08:10.000 --> 00:08:26.560
But what software has not been able to do in the past is actually extract that data from its system, do the work that your specific policies and obligations require, and then package that into the data sets or evidentiary notes that are required of your counterparty.
00:08:26.720 --> 00:08:35.200
And so what we found as we entered this landscape, right, is that by pulling on this thread of man, compliance is really reactive.
00:08:35.360 --> 00:08:39.360
How can we change this dynamic by sitting within the systems work happens?
00:08:39.600 --> 00:08:44.480
What we realize is that we were actually setting up a system to do the work ourselves.
00:08:44.639 --> 00:09:00.559
And so the big, you know, unlock for us, and I think what is really unique about Leela is that software systems in the past have typically been able to consolidate data in a way that helps people do their jobs differently, augments the experience of doing their job, et cetera.
00:09:00.639 --> 00:09:05.279
But what Leela can do is actually sit in parity with your team.
00:09:05.519 --> 00:09:08.240
We can operate like a force multiplier, right?
00:09:08.320 --> 00:09:14.960
It's like you now have eight new compliance analysts that can actually do the work across that entire chain.
00:09:15.200 --> 00:09:23.840
And we think that is sort of the big pull and interesting draw we're getting from our initial customers is man, it's not just that you're making my life more efficient.
00:09:24.080 --> 00:09:33.039
I can now meaningfully cover more ground, which we think is really important in the age of AI where agents are actually starting to do the work themselves.
00:09:33.279 --> 00:09:33.519
Yeah.
00:09:33.679 --> 00:09:36.559
I also wanted to ask more about the age of AI and compliance.
00:09:36.639 --> 00:09:41.120
Like what is happening on product teams and compliance teams as products are getting built?
00:09:41.200 --> 00:09:48.240
Because it's kind of always been a struggle to get things approved before they go live and with just with the iterative nature of uh software.
00:09:48.399 --> 00:09:54.799
And I imagine as software's changing so drastically, that's also, you know, changing how compliance is working with that.
00:09:55.279 --> 00:09:55.840
Absolutely.
00:09:56.000 --> 00:10:00.480
So it's I'll give you kind of two anecdotes that I think are um really salient here.
00:10:00.720 --> 00:10:10.240
The first is if you think about kind of the macro of how quickly engineering and product and sort of commercial shipping teams are going, right?
00:10:10.320 --> 00:10:17.039
Like I was just at a conference recently, and the entire conference floor was about that for the specific industry, right?
00:10:17.120 --> 00:10:20.320
Like, how can you go faster in your in your commercial side?
00:10:20.480 --> 00:10:22.960
How can you optimize operations, et cetera?
00:10:23.120 --> 00:10:25.120
And we had our little booth on the side, right?
00:10:25.200 --> 00:10:32.879
And the other people that were coming to our booth, what we would say to them is we give you the ability to buy, use, and procure all those products, right?
00:10:33.039 --> 00:10:38.320
Like to be able to use all of the great stuff that's out there, especially if you're regulated.
00:10:38.480 --> 00:10:45.360
It is a non-starter unless you can confirm that all of that stuff is sitting in parity with your compliance obligations.
00:10:45.600 --> 00:11:03.200
And so I think from a macro perspective, the pace at which software, product, and like AI native services are being developed on the commercial side need companion certainty, right, on the trust and risk mitigation and compliance side, which is something really interesting that we're seeing.
00:11:03.440 --> 00:11:14.480
In a more perhaps morbid example, one of the things that we talk about a lot is what happens when a loan underwriter uses AI to actually make underwriting decisions, right?
00:11:14.559 --> 00:11:18.480
Which is happening a lot and is continuing to happen more and more over time.
00:11:18.639 --> 00:11:27.440
But what if that risk decision misses something, makes a underwriting decision based on an illegal facet of a consumer that they were not allowed to, right?
00:11:27.600 --> 00:11:37.200
But instead of stopping like a human might to get approval for every single one, an agent will just compound that decision across 200 data points, right?
00:11:37.360 --> 00:11:41.120
And at the time you see it, then you're like, you know, everything is going wild.
00:11:41.279 --> 00:12:02.159
And so companies then have to make a decision now in regulated finance, particularly on whether they are gonna open themselves up to this aperture of risk that they, you know, no one would be comfortable opening themselves up to, or whether they are gonna roadblock product development by just shutting everything down until manual compliance can get through everything, right?
00:12:02.320 --> 00:12:18.240
And that feels like such a, you know, Hobson's choice for companies that we think we can make, you know, real progress in helping get them to parity, especially as so much new development kind of, you know, opens itself up to these businesses.
00:12:18.480 --> 00:12:18.879
Totally.
00:12:19.120 --> 00:12:19.679
What happens?
00:12:19.759 --> 00:12:25.759
Like, what are some examples, I guess, of like what happens if like in the fintech industry um they do fall out of compliance?
00:12:25.919 --> 00:12:29.519
Like I imagine it's a a pretty big deal, right?
00:12:29.919 --> 00:12:30.159
Yes.
00:12:30.320 --> 00:12:33.039
So there are a couple of sort of iterations of this, right?
00:12:33.200 --> 00:12:50.080
Many fintechs will work with counterparties, perhaps sponsor banks, perhaps, you know, bank as a service companies, operational counterparties, credit facilities, credit providers, you know, these kinds of folks that give them obligations or pass through obligations to the fintech.
00:12:50.399 --> 00:12:58.159
The worst case scenario we've seen a couple of times in the past, the canonical example over the last several years was Synapse and Evolve.
00:12:58.399 --> 00:13:01.679
Synapse was a fintech company, Evolve was their bank partner.
00:13:01.919 --> 00:13:15.120
And because of Synapse's mismanagement, not only did the company ultimately go under, Evolve was tagged with many, many, many regulatory challenges and issues from their regulators, right?
00:13:15.279 --> 00:13:21.759
The OCC, FINRA, larger federal rule sets have been developed specifically off of this example.
00:13:21.919 --> 00:13:28.799
And I think that is like kind of the extreme situation of what we see every day in regulated finance, particularly, right?
00:13:29.200 --> 00:13:35.200
That if you don't follow the rules, sometimes it's, you know, we can, we can figure it out afterwards.
00:13:35.279 --> 00:13:36.799
We can ask for forgiveness later.
00:13:37.039 --> 00:13:39.519
Most times it's death by a thousand cuts.
00:13:39.679 --> 00:13:42.480
You're missing so many things over time, right?
00:13:42.559 --> 00:13:45.039
That when you catch them, it's too late.
00:13:45.200 --> 00:13:48.320
And then again, you kind of apply AI on top of that, right?
00:13:48.399 --> 00:13:53.120
And the rate at which you're missing things, that risk just opens up so much more.
00:13:53.279 --> 00:14:05.679
And so I think the the real scare scenario is, man, we've missed so many things that by the time we find out about them, we're actually going to be tagged with millions of dollars in fines and fees, right?
00:14:05.759 --> 00:14:06.240
All of this.
00:14:06.480 --> 00:14:06.559
Okay.
00:14:06.879 --> 00:14:07.200
Right.
00:14:07.360 --> 00:14:07.840
Millions.
00:14:07.919 --> 00:14:09.279
I I should have mentioned that before.
00:14:09.360 --> 00:14:09.679
Yeah.
00:14:09.759 --> 00:14:18.639
Like the the aperture of like, you know, getting this wrong or the the risk of getting this wrong, rather, is quite a scary risk for many of these businesses.
00:14:19.120 --> 00:14:20.799
Yeah, for sure.
00:14:20.960 --> 00:14:25.519
I mean, like millions of dollars in fines, especially for some fintechs that are like still growing.
00:14:25.919 --> 00:14:26.720
Exactly.
00:14:26.960 --> 00:14:40.559
And then the even if you're doing everything right, right, like the the other side is that you're spending so much on making sure you're doing everything right that ultimately, even if the risk never materializes, the cost continues to grow.
00:14:40.720 --> 00:14:40.879
Right.
00:14:41.120 --> 00:14:45.679
So again, it's like this really hard place for companies to be in.
00:14:45.840 --> 00:14:48.080
Like to go back to that first principle, right?
00:14:48.320 --> 00:14:51.360
Compliance and legal only exists to help these businesses grow.
00:14:51.519 --> 00:14:59.279
And that feels very inauthentic if you're scaling nonlinear, but your compliance function has to scale linearly, right?
00:14:59.360 --> 00:15:00.960
And your cost is scaling linearly.
00:15:01.039 --> 00:15:13.840
Like that that doesn't feel and that isn't economically, you know, incentivizing businesses to invest more in these spaces, which, you know, we think that there's a big opportunity to get a ton of new value from these spaces.
00:15:15.039 --> 00:15:25.600
And just to give us a sense, like for anybody that hasn't seen the Leela product or done a demo, what does it feel like to just, you know, have Leela inside of your company doing this work?
00:15:26.159 --> 00:15:34.720
You know, is it like is it just happening in the background or is it something that people need to log into and they're, you know, kind of holding the reins, or how does it work?
00:15:35.360 --> 00:15:37.200
I I love this question, Amber.
00:15:37.519 --> 00:15:42.960
Thank you for asking this because it was a very deliberate set of choices from our team in designing the product.
00:15:43.440 --> 00:15:52.000
So, first and foremost, our goal here was you shouldn't have to do anything to have Leela add massive value to your business, right?
00:15:52.159 --> 00:15:55.279
What does that look like in the era of AI native services?
00:15:55.679 --> 00:15:59.039
We think it looks really different from the way that SaaS existed.
00:15:59.279 --> 00:16:13.759
SaaS enabled us to, again, you know, kind of harness vast quantities of data, present it in really valuable ways, and then have that presentation be the jumping off point for the user to do things in a platform or in an app.
00:16:14.000 --> 00:16:21.840
What we believe the value of Leela should be, and this is kind of how we've designed the product end to end, is work should just be happening.
00:16:22.000 --> 00:16:31.840
And then when you come into the platform, you are able to use that work or jump right in where the work left off in order to take advantage of what's gotten done.
00:16:32.080 --> 00:16:36.159
And so to answer your question very directly, Leela is always working, right?
00:16:36.240 --> 00:16:46.960
It's always doing things, searching for things, ensuring that it knows, you know, I should say maybe she knows and and really personify our AI native service here, what's going on.
00:16:47.120 --> 00:16:54.799
Now, the experience as a user then is that when you jump in, you really look at our system as four prongs, right?
00:16:54.960 --> 00:16:55.840
What is my program?
00:16:56.000 --> 00:16:57.200
What are my obligations?
00:16:57.360 --> 00:16:58.639
What work do I have to do?
00:16:58.799 --> 00:17:04.079
Where Leela left off, and where's the packaged evidence that I can download and send to my counterparty?
00:17:04.240 --> 00:17:10.559
Those four components of compliance that we described previously, Leela handles 85% of.
00:17:10.720 --> 00:17:23.599
Now, I would be remiss if it died if I did not mention, right, especially when it comes to doing work and actually making decisions, Leela will always take you up to the point that you that human judgment is required.
00:17:24.000 --> 00:17:29.519
I think one thing that is really important in spaces like legal and compliance is we can do a lot.
00:17:29.920 --> 00:17:37.279
We can, you know, of course, do more in terms of data collection that we can then present in a valuable way and do work based off of it.
00:17:37.440 --> 00:17:44.240
But the humans who actually are the experts in this space are always going to be the ones that are in control, right?
00:17:44.480 --> 00:17:59.519
And so what we've also embedded in our system is a really easy way to be able to tell what judgment is required when, and to give you the information that you need very deliberately and specifically to make that part of your life easier, right?
00:17:59.680 --> 00:18:16.000
To make it not such that you're searching across 17 different systems to actually make the decision in front of you, but rather that we've made the decision, that we've made a recommendation, given you the data to decide whether you agree with our recommendation, and then, you know, kind of be off to the races.
00:18:16.160 --> 00:18:32.160
So the way the product works is really to do all of the things that, you know, are hard about compliance while you're sleeping, so that when you come into it, you're doing only what is required of you in a way that feels like you're talking to, you know, an employee or a person on your team.
00:18:32.240 --> 00:18:35.680
And so I'm excited for you guys to use it and and and see if that's the way that it feels.
00:18:35.920 --> 00:18:40.240
Honestly, as you're talking about regulated industries and compliance, I'm like, we need this at the council.
00:18:40.400 --> 00:18:44.240
I feel like all, you know, all financial institutions need to be using Leela.
00:18:44.480 --> 00:18:45.839
So my God, it's so true.
00:18:46.079 --> 00:18:47.680
We should be testing this out too.
00:18:47.839 --> 00:18:58.799
Um you're talking about like the judgment though, it's so interesting how much we see this like across a lot of the industries that we're working in, like on the legal tech side.
00:18:58.880 --> 00:19:00.799
I think about this a lot on the health tech side.
00:19:00.960 --> 00:19:05.039
Also, there's a huge debate around, you know, will AI replace doctors?
00:19:05.119 --> 00:19:16.000
And it's like, no, AI can like really help with the clinical decision support, but doctors are still ultimately ultimately making the treatment decisions right now and using their judgment based on the information.
00:19:16.079 --> 00:19:20.720
So I see a lot of parallels in how you're thinking about this for um compliance and Leela.
00:19:21.440 --> 00:19:22.079
100%.
00:19:22.559 --> 00:19:32.480
And I think that like there in every industry, right, there is so much work that like I remember when I was a lawyer, I was like, can someone just do this annoying stuff for me so I can do the interesting stuff, right?
00:19:32.640 --> 00:19:33.920
Right and make the decision.
00:19:34.480 --> 00:19:35.200
Yeah, yeah.
00:19:35.359 --> 00:19:46.240
And I think like one thing, especially in compliance, is because so much of it is related to information gathering and also information that tends to be challenging for humans to manually gather.
00:19:46.480 --> 00:19:51.920
The fact that we can actually exceed human capacity in a certain regard, right?
00:19:52.079 --> 00:19:59.440
That we can sit in systems and watch the work happen in real time gives humans the ability to do even more, right?
00:19:59.599 --> 00:20:13.519
And I think that's a really cool feature is like maybe actually the way in which compliance will evolve as a job is much more interesting, strategic, based on the ROI that they are delivering today, but in a much clearer way, right?
00:20:13.599 --> 00:20:25.680
Like I think there's a lot of cool stuff that can come from having a more powerful system than humans maybe could design in like certain regards to unlock the things that humans don't get to do as much of in this area.
00:20:26.079 --> 00:20:29.519
You're also developing a proprietary language for compliance.
00:20:29.599 --> 00:20:31.519
Can you tell us a little bit more about reggling?
00:20:31.839 --> 00:20:32.240
Yes.
00:20:32.559 --> 00:20:38.240
Thank you again for asking that because again, very related to what we were talking about, this judgment piece, right?
00:20:38.480 --> 00:20:48.880
I think sometimes one thing that we internalized really deeply was how is it possible for a system to not exist like this yet if someone could build this in Claude?
00:20:49.119 --> 00:21:04.240
And what we ultimately realized, right, is that the fallibility of foundation models and other, you know, systems to be able to solve this problem is that you can run the same thing through a foundation model three times and you will get a different result each time, right?
00:21:04.400 --> 00:21:05.680
It will use its judgment.
00:21:05.839 --> 00:21:13.759
I mean, what is, again, a foundation model in this regard, other than a pattern recognition system based on all of the information available to it at any given time?
00:21:14.000 --> 00:21:26.079
If you apply that to a compliance problem, you won't be able to have the rate of repeatable, predictable accuracy that is required of a regulator, of an auditor, of a counterparty, right?
00:21:26.559 --> 00:21:31.680
To be able to really effectively say, I can outsource this to AI, right?
00:21:31.839 --> 00:21:44.400
And so the problem we were looking to solve was how do we build a system that's comprehensive and solves this problem, but doesn't do it in a way that will expose not only hallucinations, but also fallibility in the result, right?
00:21:44.559 --> 00:21:46.319
Or variance in the result.
00:21:46.640 --> 00:21:59.759
And our solution to that problem was to design a proprietary coding language that our agents and our AI operates both as a harness itself as well as a way in which if a, you know.
00:22:00.079 --> 00:22:01.119
We have a double harness, sort of.
00:22:01.200 --> 00:22:09.279
We've developed our own harness and benchmark system internally, but then we also have reguling, which enables us to meet that accuracy bar.
00:22:09.440 --> 00:22:25.759
What it actually is is essentially a distillation of if I were a regulator for FINRA or the OCC or the SEC or if I worked at your sponsor bank or your credit provider, so on and so forth, how would I approach this decision?
00:22:25.920 --> 00:22:32.720
How do I know what matters in MTL regulations broadly versus in lending regulations?
00:22:33.039 --> 00:22:53.680
The way we design that was with our own team of really, really amazing advisors who are ex-regulators, ex-operators, CCOs, and GCs at major institutions who have themselves contended with these problems and who have enabled us to translate all of that expertise into a regulator side language.
00:22:53.839 --> 00:23:03.119
Now we've also embedded in Regling, now that we know what to review on the regulator regulatory side, how do we know what to review on the operating side, right?
00:23:03.279 --> 00:23:10.880
If we're in Figma, how do we know what triggers a lending clause versus if we're in Google Docs?
00:23:11.279 --> 00:23:19.279
And so the other part of Reglang was sort of translating all of that expertise into the way work actually happens in the systems that we work with.
00:23:19.519 --> 00:23:27.519
And the more that we embed, right, and integrate and build, the deeper, more entrenched and more sophisticated Reglang becomes.
00:23:27.839 --> 00:23:40.799
And so that's been a big reason I think we've been, you know, successful with our clients to date and continue to be able to serve clients as we expand with that level of accuracy, which we think is just non-negotiable in this space.
00:23:41.039 --> 00:23:41.359
Yep.
00:23:41.599 --> 00:23:42.240
Absolutely.
00:23:42.480 --> 00:23:48.799
And can you tell us a little bit more about the type of customers you're trying to serve now in the future and who's already working with Leela if you can?
00:23:49.119 --> 00:23:49.440
Yes.
00:23:50.160 --> 00:23:57.279
So we are primarily poised today to work with anyone who would call themselves a financial services company, right?
00:23:57.359 --> 00:24:02.799
So on one side, um, you know, banks, funds, traditional financial services.
00:24:03.039 --> 00:24:19.680
On the other, um, insurance companies, payments companies, lending companies, crypto companies, anyone who sort of has a ton of stuff they need to manage from a regulatory perspective that needs to be managed in various, like very small obligations across their business, right?
00:24:20.079 --> 00:24:38.880
And so what that has enabled us to kind of penetrate so far, we've gone all the way from amazing design partners like built and crypto.com, these big enterprise companies that have, you know, various needs that we're able to serve, all the way to very small startups who don't even know where to start with compliance.
00:24:39.039 --> 00:24:42.240
And because we can take on all of the surface area we can, right?
00:24:42.400 --> 00:24:54.960
Policies, obligations, work, and evidence that enables us to essentially run their compliance function alongside their fractional CCO or their full-time CCO, um, or whomever they want to make the decisions, right?
00:24:55.039 --> 00:25:02.000
So we've been lucky enough to kind of span that core size chasm so far.
00:25:02.240 --> 00:25:11.759
And we think that frankly, anyone who has a lot of, you know, regulatory obligations that require them to do work to uphold them is a good fit for Leela.
00:25:12.079 --> 00:25:12.400
That's awesome.
00:25:12.559 --> 00:25:19.200
And I imagine it's a fairly like global customer base too, just because, you know, a lot of these companies, just in finance, they operate globally.
00:25:19.279 --> 00:25:22.400
So just being able to handle compliance, you know, in that sphere.
00:25:22.480 --> 00:25:24.559
I'm curious if that's relevant for you guys.
00:25:24.799 --> 00:25:25.279
Yes.
00:25:25.519 --> 00:25:33.759
I think one thing that's really cool about running this business is, you know, in my old roles, international expansion was always like such a big question, right?
00:25:33.839 --> 00:25:34.960
Like when do we do it?
00:25:35.119 --> 00:25:37.119
How do we think about doing it, etc.?
00:25:37.440 --> 00:25:49.440
And it's largely because, again, with software, the way in which you, you know, roll out tends to be, you know, you have to follow X amount of rules and then productize that rule set in various different regimes, right?
00:25:49.599 --> 00:25:52.960
Especially in in kind of semi-regulated or regulated industries.
00:25:53.200 --> 00:26:05.119
Now, because we're able to ingest all of that and essentially underwrite it as we expand quickly and easily, especially with the addition of Regling, it's kind of agnostic of where you operate.
00:26:05.359 --> 00:26:09.440
So, yeah, like one of the companies we're talking to right now is based in the UK.
00:26:09.599 --> 00:26:12.799
There's another person that we're talking to that's based in Dubai, right?
00:26:13.039 --> 00:26:26.799
I think especially, you know, as regimes sort of evolve across the country, it is so hard to figure out how to manage your operations to match those, you know, varying colorful regulations that you need to follow.
00:26:27.039 --> 00:26:37.839
And exactly as you said, Amber, because our system is so agile, it allows us to be able to sort of, you know, penetrate the way in which that flow works no matter where you operate.
00:26:37.920 --> 00:26:40.960
Um so we'll kind of see how that evolves.
00:26:41.039 --> 00:26:47.759
Um, we aren't, you know, all of that to say there's no reason for us today to be particularly focused on a geography.
00:26:47.920 --> 00:26:55.920
Um, and and you know, we think that that enables customers who may not have been served by traditional products to potentially be able to work with us as well.
00:26:56.079 --> 00:26:56.319
Yeah.
00:26:56.559 --> 00:27:06.000
And then are you staying like within the FinTech or financial services space, or do you see potential for Leela to even go into other industries, regulated industries beyond that?
00:27:06.319 --> 00:27:07.279
I love this question.
00:27:07.519 --> 00:27:11.519
Today we are very focused on financial services for a couple of reasons.
00:27:11.759 --> 00:27:20.079
Reglang is excellent in financial services and the way in which we are able to approach the breakdown of the way in which regulated finance works.
00:27:20.319 --> 00:27:23.200
Again, and we have a pretty broad definition of finance, right?
00:27:23.359 --> 00:27:28.559
Insurance and payments and crypto and trad fi are sort of all in that bucket.
00:27:28.880 --> 00:27:30.240
It's a lot to tackle on its own.
00:27:30.480 --> 00:27:37.920
Yeah, and we believe that we can do it because there are broad patterns across the way in which those businesses operate their compliance surface area.
00:27:38.160 --> 00:27:58.559
Now, one thing that I think is really fascinating about your question, and one thing we're actively thinking about is when I talk to my friends about what we're building, if they've been in healthcare or beauty or food, right, in other areas where regulation is still super, super critical and high and important, they can very easily see the application of this and in their lives as well.
00:27:58.799 --> 00:28:06.079
And so our vision for Leela is ultimately that this can just be the way that regulated compliance is run, right?
00:28:06.240 --> 00:28:22.960
That you can actually embed Leela within the way in which your compliance operators are working, regardless of where you are in a regulated business, and that ultimately this becomes a reliable way for you to scale up that function, scale up your business.
00:28:23.119 --> 00:28:31.680
There is nothing novel about the way in which the technology can be applied, other than the fact that reguling needs to get a little bit more sophisticated, right?
00:28:31.839 --> 00:28:38.559
And as that scales, we believe we really do have an opportunity to change the way in which regulated compliance works altogether.
00:28:38.799 --> 00:28:39.119
Awesome.
00:28:39.279 --> 00:28:39.599
Awesome.
00:28:39.759 --> 00:28:41.599
And we would love to hear a little bit more about the team.
00:28:41.680 --> 00:28:44.240
Can you tell us about your co-founder and how you two met?
00:28:44.400 --> 00:28:46.720
And I'd love to hear more about that dynamic.
00:28:46.960 --> 00:28:47.359
Oh my God.
00:28:47.440 --> 00:28:52.799
I'm really excited to talk about this because I think when I was starting a business, this was like so myopic to me.
00:28:52.880 --> 00:28:54.559
Like, how do people meet their co-founders?
00:28:54.640 --> 00:28:56.720
Like totally everybody has so much pressure.
00:28:56.799 --> 00:28:59.279
Like, oh my God, like where do you find that person?
00:28:59.599 --> 00:28:59.839
Yes.
00:29:00.000 --> 00:29:03.200
And I think especially because of the roles that I was in before, right?
00:29:03.359 --> 00:29:11.440
Like I had the experience of like building a scrappy team from the ground up, but but building with a partner is so different.
00:29:11.519 --> 00:29:14.480
Like I hadn't had that experience before, right?
00:29:14.559 --> 00:29:16.880
And so I so I love talking about this.
00:29:17.039 --> 00:29:20.000
So my co-founder is named Frank Kim.
00:29:20.160 --> 00:29:24.960
He is a two-time founding CTO before Leela.
00:29:25.279 --> 00:29:31.200
And and so, you know, I I have a lot that I'm learning from him as well because he's seen this part right before.
00:29:31.359 --> 00:29:36.400
He's built video games in the past, he's been sports uh built sports data analytics systems.
00:29:36.559 --> 00:29:40.640
He's really kind of an engineer's engineer.
00:29:40.799 --> 00:29:43.839
He's incredible at building things that people want.
00:29:44.079 --> 00:29:47.200
I met him actually through LinkedIn.
00:29:47.440 --> 00:29:47.599
Wow.
00:29:47.920 --> 00:29:54.559
So I spent some yes, so I spent some time when I uh knew I wanted to start a business thinking about like who I wanted to do it with.
00:29:54.720 --> 00:30:02.480
I have an amazing network of amazing operators, but none that I could quite think about, you know, partnering with for this like tenure journey.
00:30:02.640 --> 00:30:12.559
And so I spammed a bunch of people on LinkedIn that I admired, that I saw, you know, their backgrounds and thought to myself, man, if I could really, you know, if I had the opportunity to work with this person, I know I'd be better.
00:30:12.720 --> 00:30:15.759
Hank luckily somehow responded to me.
00:30:16.079 --> 00:30:26.079
And so we spent a couple of months, yeah, like iterating together, seeing what it would be like to work together and then committed and have kind of been working together ever since.
00:30:26.319 --> 00:30:30.160
We recently had two new incredible ads to our team.
00:30:30.400 --> 00:30:37.759
One engineer who before this founded a competitor to N8N that he open sourced and you can, I think, still use it.
00:30:37.839 --> 00:30:40.319
It's called Hot Dev if anyone wants to check it out.
00:30:40.559 --> 00:30:42.000
His name is Curtis Summers.
00:30:42.160 --> 00:30:56.240
He has enabled us to really 1000 X our engineering function and built product that I think, frankly, it would have taken me like five years to build before we built in like two months, which is which is it's incredible.
00:30:56.400 --> 00:30:59.119
Like, I cannot, I mean, we should have a separate sesh on this.
00:30:59.200 --> 00:31:03.839
Like, it is so cool what you can do with amazing talent and AI today.
00:31:04.160 --> 00:31:17.039
And then we have an incredible growth associate who has really internally up-leveled us in terms of how we rely on AI, how we've created internal tooling for ourselves that then actually support our customers.
00:31:17.279 --> 00:31:25.920
Her name is Tay Zakowski, and she's just, you know, really rounded out for us what we think is our all-star team as we start out.
00:31:26.079 --> 00:31:38.480
Um, I think for me, you know, if if if if I may, the thing that has made everything just so incredible for us to date are the amazing customers that we've been able to learn from and that we've been able to serve.
00:31:38.640 --> 00:31:45.440
Um, I think we improve on a daily basis because of the ability that we have to serve these incredible users.
00:31:45.599 --> 00:31:47.519
Um, and then my amazing team.
00:31:47.680 --> 00:31:49.839
Like, I can't imagine doing this without them.
00:31:50.000 --> 00:31:53.359
It's made it simultaneously the most fun ever.
00:31:53.519 --> 00:32:00.559
And also a problem that I feel like we can actually tackle because I'm surrounded by all-star players, right?
00:32:00.640 --> 00:32:02.400
So, how how could anything go wrong?
00:32:02.559 --> 00:32:04.480
I mean, talk to me in six months and I'll tell you.
00:32:04.559 --> 00:32:06.400
But but it's been incredible so far.
00:32:06.720 --> 00:32:07.279
It's awesome.
00:32:07.440 --> 00:32:14.559
And I imagine like some of the team growth has to do with you're just able to raise millions of dollars toward toward your vision for this company.
00:32:14.720 --> 00:32:21.359
And so, can you tell us about like the decision to raise venture capital and the process of raising venture capital and what you plan to do with it?
00:32:21.680 --> 00:32:22.319
Yes.
00:32:22.640 --> 00:32:23.119
Yes.
00:32:23.359 --> 00:32:33.599
So it was quite a I I I have been lucky enough, I'll put it that way, to partner with truly incredible investors on this journey so far.
00:32:33.759 --> 00:32:35.599
You guys are obviously two of them.
00:32:35.759 --> 00:32:54.960
So, yes, we raised$4.6 million from Array Ventures, Female Founders Fund, the Council, Northwestern Mutual, Future Ventures, and then an array, no pun intended, of incredible financial services, um, angel investors, folks from Chime and JP Morgan and Daily Pay, and you know, more and more.
00:32:55.039 --> 00:32:55.920
The list goes on.
00:32:56.160 --> 00:33:05.519
The reason that we wanted to raise venture capital is the product that we are building is a very challenging one and one that I think can impact millions of people.
00:33:05.759 --> 00:33:18.240
I really authentically, in my heart of hearts, believe that this is the way that compliance should be run in regulated financial services and beyond, to be able to eat into that market, right, at the rate that we want to.
00:33:18.319 --> 00:33:21.200
And that at the rate, I think demand will scale.
00:33:21.519 --> 00:33:29.279
It was, I think, a really obvious decision at this stage that we wanted to be able to really up level, you know, very quickly.
00:33:29.440 --> 00:33:39.039
And I think this capital raise allows us to test, iterate, put more product down and scale at the rate that, you know, we think that that we should and that we need to.
00:33:39.200 --> 00:33:47.680
Now, on the process side, I think the market has changed so, so much in the last couple of years, which was really interesting for me to navigate.
00:33:47.920 --> 00:33:52.319
We have been a part of capital raising processes in my past life many, many times.
00:33:52.480 --> 00:34:07.039
And I think what's different now is that the story, the narrative, the way in which you can actually pinpoint why you are the right person to solve a problem, not just because I have the background or whatever, but that you see something that others don't.
00:34:07.200 --> 00:34:08.480
That's always been important.
00:34:08.639 --> 00:34:16.159
But I think that it is more important than ever now, not just for the storytelling side, but so that you can move as quickly as you need to, right?
00:34:16.320 --> 00:34:21.760
Like I don't I don't think it is enough to say anymore, right?
00:34:21.840 --> 00:34:25.360
Like, I think this is the case and I want to run an experiment around it.
00:34:25.599 --> 00:34:28.000
You have to have that depth of conviction.
00:34:28.239 --> 00:34:35.440
Um, and I think that really served me in this process was doing that work ahead of, you know, going out to raise.
00:34:35.519 --> 00:34:48.880
So I think if I had one piece of advice to any other first builders who know they have, you know, an inkling around a problem they want to solve, a lot of stuff that they, you know, think they know that's special and that no one else does is get started, right?
00:34:49.039 --> 00:34:55.840
Like start doing stuff, start testing that thesis, show that that actually is or is not what people want.
00:34:56.079 --> 00:35:15.280
And as you start building your own conviction, I think the capital raising process, especially at like pre-seed and seed, right, will become not only much easier, but you'll be able to surround yourself with people who actually see and amplify that conviction, which I think has been really, really critical for us in these early days.
00:35:15.440 --> 00:35:15.599
Yeah.
00:35:15.840 --> 00:35:18.480
And with AI right now, I feel like anybody can build anything.
00:35:18.559 --> 00:35:20.800
And so there's a lot of noise out there for investors too.
00:35:20.960 --> 00:35:24.159
Like, yes, you know, there's probably so many, like name any topic.
00:35:24.239 --> 00:35:26.719
There's an AI app layer thing for it.
00:35:26.960 --> 00:35:36.880
And so it's if you can really like get in the room, accept people with an insight, explain your conviction, and like actually have the conviction that people feel in the room, that goes such a long way.
00:35:36.960 --> 00:35:38.320
And we definitely felt that with you.
00:35:38.559 --> 00:35:38.880
Yeah.
00:35:39.119 --> 00:35:40.000
Thank you so much.
00:35:40.239 --> 00:35:41.280
100% though.
00:35:41.519 --> 00:35:51.280
I remember when Amber actually like was telling me about Leela and meeting you, and just like the way that she was able to explain it, you know, so concisely and the problem with the floor.
00:35:51.440 --> 00:35:53.119
Which is wild, we're taking compliance.
00:35:53.199 --> 00:35:54.960
I'm like, I love it.
00:35:55.199 --> 00:35:55.840
I love it.
00:35:56.079 --> 00:35:57.519
Oh my God, amazing.
00:35:57.679 --> 00:35:59.119
No, it's so true though, right?
00:35:59.199 --> 00:36:03.440
Like, I think that there are so many complicated things that AI can unlock.
00:36:03.679 --> 00:36:14.639
And those builders that are able to kind of pinpoint why a certain thing is the thing that will unlock that industry or that complicated workflow, hopefully are gonna be the ones that that win.
00:36:14.719 --> 00:36:20.159
But I love that story because that makes me feel like, okay, I gotta go back to my notes and make sure I'm saying that over and over again, right?
00:36:20.239 --> 00:36:23.440
Like to make sure that people actually come back like Amber did.
00:36:23.519 --> 00:36:24.480
So I'm happy to hear that.
00:36:24.559 --> 00:36:30.639
I'm glad that I'm glad that it um made sense to you because it seems like it worked for a lot of others too, because it was a successful raise.
00:36:30.719 --> 00:36:34.480
So I think you ended up raising a lot more than you had even set up to raise, which is awesome.
00:36:34.559 --> 00:36:34.639
Yes.
00:36:34.800 --> 00:36:37.280
So we were lucky enough to be oversubscribed.
00:36:37.360 --> 00:36:44.800
And I think that, you know, one thing I will also say, just to close out on the on the venture side, is I I hope that other founders are like not undeterred.
00:36:44.960 --> 00:36:52.559
I think a lot of times when I hear about raising and sort of what people are building, it is simultaneously inspiring and can be intimidating.
00:36:52.639 --> 00:36:58.559
And I think the thing that I will say is like what has served us so far is just like just keep going, you know?
00:36:58.719 --> 00:37:04.159
Like I if you believe something that exists, yeah, be get in there one thing.
00:37:05.840 --> 00:37:06.400
Exactly.
00:37:06.880 --> 00:37:07.840
Exactly.
00:37:08.239 --> 00:37:08.639
Yeah.
00:37:08.880 --> 00:37:18.320
And I know it's early, but any learning so far in like how Leela has been used or implemented by like your design partners or customers and anything you're super excited to see next?
00:37:18.639 --> 00:37:20.400
I I've said this to every question.
00:37:20.559 --> 00:37:21.599
I love that question too.
00:37:21.760 --> 00:37:23.280
Um so I will leave you guys with this.
00:37:23.440 --> 00:37:27.360
When we initially started Leela, our thesis was that compliance is reactive.
00:37:27.519 --> 00:37:28.480
We can change that, right?
00:37:28.559 --> 00:37:33.519
We can sit within systems, we can actually see what's happening and we can help transform the function.
00:37:33.760 --> 00:37:42.800
But as we did that for our design partners, for our initial paying customers, what we realized is that what we actually were unlocking was the capacity to do work, right?
00:37:42.960 --> 00:37:46.800
And that is really the dynamic game changer in this space.
00:37:47.039 --> 00:38:05.760
Never before has a system been able to actually derive data and then do things with that data natively within the systems that you are working in, in order for you to package all of your bespoke obligations, not, you know, ones that that are standardized, but bespoke obligations in a reliable way.
00:38:05.920 --> 00:38:12.719
And so I think that the thing that was most surprising to us and that we're most excited about is how much more work can we do for you, right?
00:38:12.960 --> 00:38:25.519
Now that we've given you and and developed, you know, for you, customer, a system that basically unlocks a lot of day one value in actually getting stuff done, what does that expand into, right?
00:38:25.599 --> 00:38:34.400
How can we go even a layer deeper in terms of the infrastructure stack and and solve problems perhaps before they become problems, right?
00:38:34.559 --> 00:38:39.039
Um, I think when I started out, I had this tagline of prevent violations before they happen.
00:38:39.199 --> 00:38:48.480
And what we're realizing is even in, you know, the regard that we exist today, sometimes things don't even get to violation because we're actually doing the work that prevents that, right?
00:38:48.639 --> 00:38:56.000
But imagine that that infrastructure layer sits even, you know, one level deeper in your data systems, perhaps, right?
00:38:56.239 --> 00:39:00.159
Or alongside the LLMs that you're utilizing to do more work.
00:39:00.320 --> 00:39:07.280
Like maybe we can actually build APIs that imbue within, you know, the futuristic version of Figma that you use.
00:39:07.440 --> 00:39:13.119
So we think that the opportunity of actually doing work has been a big unlock for us far.
00:39:13.280 --> 00:39:31.039
And we'd like to keep, you know, kind of riding that opportunity as it evolves over the next couple of years to really expand our vision, both to become more industry agnostic, as we talked about, but also enrich the customer experience in whatever way feels authentic and native to them.
00:39:31.199 --> 00:39:34.559
So much more to come on that in the coming, you know, months and years.
00:39:34.719 --> 00:39:45.360
But I think that's been the most exciting part for us is man, in solving one problem, we also discovered that we could do so much more, you know, compra add so much more comprehensive value to our end users.
00:39:45.599 --> 00:39:46.079
That's awesome.
00:39:46.239 --> 00:39:49.920
And where can our listeners keep up with your journey at Leela and beyond?
00:39:50.559 --> 00:39:55.039
Yes, you can follow us on LinkedIn and Twitter and Instagram.
00:39:55.440 --> 00:40:00.480
On LinkedIn, our company uh it's LinkedIn.com slash company slash with Leela.
00:40:01.280 --> 00:40:03.199
On Instagram, Leela Inc.
00:40:03.440 --> 00:40:05.280
and on Twitter also Leela Inc.
00:40:05.920 --> 00:40:08.559
You can also follow me on all of those platforms.
00:40:08.880 --> 00:40:14.000
On LinkedIn, I'm Preeti Krishnan, and then Twitter and Instagram, P Krishnan.
00:40:14.079 --> 00:40:15.280
So please follow along.
00:40:15.440 --> 00:40:19.679
I'm super excited for as many people as possible to be on this ride.
00:40:19.920 --> 00:40:28.480
If you'd like to try Leela, we also have an interactive demo experience that you can access on try dotwithle.com.
00:40:28.880 --> 00:40:32.880
And then uh you can also learn more about us on our website with leela.com.
00:40:32.960 --> 00:40:34.639
So those are a lot of hyperlinks.
00:40:34.800 --> 00:40:37.760
Um, but please feel free to them all in the show notes for sure.
00:40:38.079 --> 00:40:38.559
Perfect.
00:40:38.719 --> 00:40:39.440
Perfect.
00:40:39.760 --> 00:40:40.480
Amazing.
00:40:40.880 --> 00:40:42.639
Well, thank you so much for coming on the show.
00:40:42.800 --> 00:40:43.360
That was great.
00:40:43.599 --> 00:40:43.679
Yeah.
00:40:44.079 --> 00:40:45.360
Thank you both for having me.
00:40:45.519 --> 00:40:50.639
It was so much fun and thank you for being such a special part of building the company to date.
00:40:50.800 --> 00:40:53.199
I'm so excited to to keep doing this with you both.
00:40:53.440 --> 00:40:54.079
We are honored.
00:40:54.159 --> 00:40:54.800
So thanks.
00:40:55.199 --> 00:40:55.840
Thanks, guys.