CodeRabbit CEO Harjot Gill on software's new bottleneck and zero to $50m ARR in 2 years

WEBVTT
00:00:00.000 --> 00:00:04.320
What A is doing is like taking the best practices it has learned to an average person.
00:00:04.480 --> 00:00:05.759
For example, I can do marketing.
00:00:05.839 --> 00:00:09.199
I may not be the best marketer, but it does level the playing field for me.
00:00:09.279 --> 00:00:15.199
So we're seeing that imbalance where sometimes you'll see that people are being held back because they're too afraid to let go.
00:00:15.359 --> 00:00:21.600
Right now, I would say is the best time I've seen in the history to do startups and create value.
00:00:23.920 --> 00:00:25.199
Hello, fellow data nerds.
00:00:25.280 --> 00:00:26.480
My guest today is Harja Gill.
00:00:26.559 --> 00:00:35.920
Harjett is the co-founder and CEO of Code Rabbit, the AI code review company that became the most installed AI app on GitHub and GitLab.
00:00:36.079 --> 00:00:41.759
Over the last two years, they've gone from zero to over 50 million ARR and over a billion valuation.
00:00:41.920 --> 00:00:43.679
Harja, welcome to Sumation.
00:00:44.159 --> 00:00:45.520
Yeah, thanks for having me here on.
00:00:46.000 --> 00:00:47.280
Really excited to be on the show.
00:00:47.520 --> 00:00:48.159
Really excited.
00:00:48.240 --> 00:00:52.560
Now just a full disclosure: Flex is an investor in CodeRabbit.
00:00:52.719 --> 00:00:53.759
We invested in the seed.
00:00:53.840 --> 00:00:55.840
We invested in the A, we invested in the B.
00:00:55.920 --> 00:00:57.200
We're about to invest in the C.
00:00:57.280 --> 00:00:58.320
We love CodeRabbit.
00:00:58.479 --> 00:01:03.679
So that's the full, you know, disclosure before we get going on the show.
00:01:04.079 --> 00:01:08.079
Now, AI is writing more code every month.
00:01:08.239 --> 00:01:15.280
And I think the bet that you were making is that the bottleneck is going to move from like writing to reviewing.
00:01:15.920 --> 00:01:17.599
How has that played out?
00:01:18.000 --> 00:01:30.239
And it played out even in more interesting ways than we imagined it would play out like two years back, especially with the capabilities of these agents improving and leapfrogging every few months.
00:01:30.799 --> 00:01:36.000
Now we're seeing like these agents are becoming more capable, able to run for longer periods of time.
00:01:36.400 --> 00:01:42.159
And recently also we are now seeing like the adoption of background agents take off.
00:01:42.239 --> 00:01:44.799
As you can imagine, like the code is getting generated.
00:01:46.000 --> 00:01:49.200
Not just from CLI and your code editors.
00:01:49.280 --> 00:01:52.000
It's being also being generated by linear.
00:01:52.480 --> 00:01:54.879
Yesterday Notion announced that they will be writing code.
00:01:55.280 --> 00:01:57.840
Datadog wants to write code, post hoc wants to write code.
00:01:58.000 --> 00:02:05.920
Almost like the floodgates opened and now you have this deluge of pull requests that are getting created, right?
00:02:06.079 --> 00:02:18.719
So the human attention is now all about how do I validate and how do I decide and build trust in whether the changes that we're seeing from these coding agents can be shipped safely or not.
00:02:19.039 --> 00:02:20.639
So yeah, I mean it's pretty interesting.
00:02:20.719 --> 00:02:24.960
Like, I mean, we when we started that time was just AI co-pilots.
00:02:25.039 --> 00:02:31.280
You were like doing tab completion, the GitHub co-pilot cursor came in a little bit around the same time as Code Rabbit started.
00:02:31.599 --> 00:02:37.360
So the velocity was very different, and the code review was still like very much human in the loop.
00:02:37.439 --> 00:02:42.479
Like humans were reading Code Rabbit comments and going and addressing those comments.
00:02:42.719 --> 00:02:46.240
But increasingly, what we are seeing is now we are part of the loop, right?
00:02:46.400 --> 00:02:56.159
So it's just agents reading the comments from this adversarial review that we provide and get to a really good implementation before some things get shipped.
00:02:56.639 --> 00:03:10.319
And it does seem like even most organizations that are of any size now, and once once you get past maybe 10, 20 engineers, they're using many different systems to write their code, right?
00:03:10.400 --> 00:03:22.159
They've got codecs, they've got Claude Code, they've got uh cursor, they've got many, many other systems that are, and as you mentioned, you know, Notion, now all these other folks are helping them write code.
00:03:22.319 --> 00:03:27.199
And so there isn't like one system that is like the main input to their code.
00:03:27.280 --> 00:03:28.240
Is that correct?
00:03:28.560 --> 00:03:29.199
That's right.
00:03:29.280 --> 00:03:40.560
I mean, this has become more of like a taste thing, just like every other tool in the developers that have in the past, as you know, like it was about Vim versus Emacs, a lot of the editor wars.
00:03:40.639 --> 00:03:48.240
Now we are seeing the harness wars or the model wars, like you have cults forming, and it's really about the the choice of the model.
00:03:48.479 --> 00:03:56.240
Like people have built fondness to opus versus pragmatic personality of GPD 5.6 or 5.5.
00:03:56.800 --> 00:04:01.520
And then it's about the surface and the harness where you want to consume these models.
00:04:01.599 --> 00:04:07.919
So it can be in the IDE, that's still like where cursor is still very popular as an IDE solution.
00:04:08.080 --> 00:04:10.719
Then in the terminal, you have cloud code CLI.
00:04:10.879 --> 00:04:19.040
And now very recently we have seen uh a lot of developers, including myself, use desktop applications where Codex Desktop app is kind of ahead of the Mac.
00:04:19.439 --> 00:04:28.160
So yeah, so it's kind of like what the organizations are now doing is like giving developers freedom on which model, harness, surface they want to consume these agents.
00:04:28.480 --> 00:04:33.839
Um, and then when it comes to code reviews, that's where you want to centralize as the guardrail.
00:04:33.920 --> 00:04:42.000
Um, while you have proliferation of these coding agents, everything has to go through a single choke point, which is now code rabbit and your GitHub and and so on.
00:04:42.079 --> 00:04:42.560
Yeah.
00:04:42.879 --> 00:04:54.079
I personally like when I even when I'm in the same repo with the, you know, let on my GitHub, I will sometimes use Codex and sometimes use Claude Code for different reasons.
00:04:54.160 --> 00:05:01.680
And I don't know, sometimes I get excited about one versus the other, or it could be like so one week I'm using one, the next week I'm using the other.
00:05:01.839 --> 00:05:08.639
And so even like one person may have like, in some ways, weird personalities and move around different systems.
00:05:09.279 --> 00:05:10.079
No, that's interesting.
00:05:10.160 --> 00:05:11.920
I we have seen that with our team as well.
00:05:12.079 --> 00:05:20.639
Like, for example, like when it comes to chick chat and like planning, like you will find cloud code models to be more uh curious and guide you in the right direction.
00:05:20.800 --> 00:05:26.879
Whereas uh the open air models are more like you just tell them what to do and then they'll no questions asked.
00:05:26.959 --> 00:05:28.319
They will just go and do it, right?
00:05:28.399 --> 00:05:29.279
What you ask.
00:05:29.519 --> 00:05:38.879
So yeah, I think people are also like seeing mixing the coding agents for various reasons because the capabilities do differ across different model providers, the personalities of these tools.
00:05:39.199 --> 00:05:48.800
Or you just run out of like credits in one and you switch to the other for the rest of the week, and then you jump to the other, and it's like you know, and it's like and they you're kind of like in there or doing it.
00:05:49.120 --> 00:05:49.680
That's right.
00:05:49.759 --> 00:06:02.319
I mean, right now we are also living in the time when OpenAI entropic are kind of testing the market on these hundred, two hundred dollar plans, which they feel that would be the most stable pricing point for these coding agents.
00:06:02.480 --> 00:06:10.240
And currently they're subsidizing, so of course there are like some rate limits, but as you know, open AI keeps resetting and and so on.
00:06:10.399 --> 00:06:22.720
And at the same time, the once you're hooked to these coding agents, like there's a lot of anxiety, like a lot of the developers are feeling so empowered that the first thing they're doing waking up is like looking at their coding agent, and that's the last thing they do when they're going to sleep.
00:06:22.800 --> 00:06:25.199
They're just keeping these systems busy.
00:06:25.680 --> 00:06:33.519
Um and uh and and that's where like a lot of people are like switching harnesses and buying multiple of these like$100,$200 plans.
00:06:33.839 --> 00:06:34.639
Yeah, yeah.
00:06:34.720 --> 00:06:37.199
Yeah, I personally have two$200 plans going.
00:06:37.360 --> 00:06:40.160
I mean, probably maybe it may add a third one at some point.
00:06:40.399 --> 00:06:42.720
Like, how does like organization think of it?
00:06:42.959 --> 00:06:55.279
Because like I imagine with all these people committing, all these people doing, you know, committing different types of code, they're they want to have some sort of better, some more standard review process, right?
00:06:55.360 --> 00:07:13.680
They would then they and each organization may want to have some sort of centralized decision of how they should do reviews, what's important to do reviews, how much should they think about security versus speed versus like how does how does one think about like the review process as an organization?
00:07:14.000 --> 00:07:17.199
Yeah, I think people are looking at like SDLC as a whole.
00:07:17.360 --> 00:07:36.240
Like, if you look at the software development lifecycle, like pre-genai, there was a very distinct planning phase because downstream your engineering resources were very expensive, so you needed to be sure that what they were building or working on at the same at any time was the right thing to work on.
00:07:36.399 --> 00:07:42.560
That's where you had linear, Jira, Notion, bunch of these planning tools come in, right?
00:07:43.040 --> 00:07:47.360
And downstream, then you had code reviews because you had the these are the outer loops, right?
00:07:47.439 --> 00:07:48.560
So they were always collaborative.
00:07:48.639 --> 00:07:55.279
So your planning was collaborative, then the developers will work on their machine, and then the reviews were always collaborative, right?
00:07:55.519 --> 00:08:08.560
So as soon as the middle inner loop became fully automated, now we can go from a prompt to an implementation which is thousands of lines within like a few minutes or hours, right?
00:08:09.360 --> 00:08:11.920
The collaboration has broken now, right?
00:08:12.079 --> 00:08:16.160
So agents are good at writing code, but um the collaboration is what suffers, right?
00:08:16.240 --> 00:08:20.000
So you have now processes broken upstream and downstream.
00:08:20.399 --> 00:08:23.759
Planning is being bypassed increasingly.
00:08:24.000 --> 00:08:34.240
So there is no reason to track something when you can just go from an idea to implementation without going through linear, and that's why linear is also saying issue systems are dead.
00:08:34.320 --> 00:08:38.080
They wrote a blog about it and they are like going downstream into code reviews themselves.
00:08:38.799 --> 00:08:43.440
And code reviews become the only choke point because really the floodgates have opened.
00:08:43.519 --> 00:08:45.919
There's no process or choke point upstream.
00:08:46.159 --> 00:08:49.840
You can't really tell these agents to get an approval before opening a PR.
00:08:49.919 --> 00:08:56.399
I mean, that work stream, that system of plans or approvals does not exist today, right?
00:08:56.639 --> 00:09:01.679
Git is the only approval-based system that we have, the workflow around that, the pull request model, right?
00:09:01.919 --> 00:09:09.679
So naturally, like everything's now hitting GitHub, and no wonder GitHub is failing to scale as well, just the PR volume has increased so much.
00:09:10.000 --> 00:09:11.120
So it's very interesting.
00:09:11.200 --> 00:09:16.720
Like the review process, when companies look at it, they clearly see that becoming a pain point.
00:09:16.799 --> 00:09:18.639
Even pre-Gen AI, it was a pain point.
00:09:18.799 --> 00:09:25.120
Around 20 to 25% time, every software engineer, like especially the senior talent, would spend on code reviews.
00:09:25.360 --> 00:09:27.200
Now it's even more than that, much more.
00:09:27.440 --> 00:09:38.480
And in fact, to a point, I'll tell you like when we enabled background agents in our code base, like a few months back, we launched a Slack agent where which could respond to Datadog alerts or support tickets.
00:09:38.799 --> 00:09:48.799
We have over 1,000 pull requests open in GitHub right now, even though we are like automating the reviews, but even deciding what to ship, what to spend time on.
00:09:48.960 --> 00:09:51.360
Like processes are broken, our CI CD was broken.
00:09:51.440 --> 00:09:55.440
So it's not just the review, even like downstream, how do you build, validate?
00:09:55.600 --> 00:10:03.039
So a lot of things suddenly became um uh hit the breaking point as the middle loop in the loop got automated.
00:10:03.360 --> 00:10:06.320
I built this like super silly app for myself.
00:10:06.480 --> 00:10:23.279
And even when I built this app, I had I I had to use, I use Claude Code, then I use Codex, I had GitHub, I had Superbase, I had Vercell, I had Modal, I had Resend, I ended up using Code Rabbit as well, I had a bunch of other tools just to build this like kind of a silly app there.
00:10:23.519 --> 00:10:27.360
Like, how do you think about going across all of these different things?
00:10:27.440 --> 00:10:34.799
Because it's similar, it's like, yeah, there's the code, which might sit on GitHub, but it's interacting with Vercel.
00:10:35.200 --> 00:10:44.399
And you have to understand what Vercel is, and then it's interacting with Superbase, and you have to understand, okay, well, what is the the good, you have to understand, like, okay, how does this work with Superbase?
00:10:44.559 --> 00:10:46.159
Because it might not work with something else.
00:10:46.320 --> 00:10:48.960
And then it's interacting with modal, which has this like.
00:10:49.039 --> 00:10:55.200
So, how do you think about like these like tool ecosystem that that kind of interacts with the code?
00:10:55.519 --> 00:10:56.639
No, that's a great question.
00:10:56.879 --> 00:11:02.720
In fact, now with the agents, you can pretty much automate a lot of the DevOps and operational PCs as well.
00:11:02.799 --> 00:11:07.120
And it's really now a matter of bringing the right tools to the agents.
00:11:07.200 --> 00:11:10.399
It's about the surface area they can access, right?
00:11:10.639 --> 00:11:17.759
And right now, if you look at the model capabilities, they're right there, but the products that capture the capabilities are still lagging behind, right?
00:11:17.919 --> 00:11:23.679
To the point that it's really the model labs pushing some of those capabilities by adding plugins in cloud code and so on.
00:11:24.000 --> 00:11:26.080
Yeah, sometimes it's like, okay, we're stopped here.
00:11:26.240 --> 00:11:36.720
You have to go into Superbase yourself and then you know put this little API key in there, go into Vercell, and it's like, oh great, like I didn't know how to go log into UI to go do this.
00:11:37.840 --> 00:11:44.240
Yeah, providing access to these agents is the challenge right now that everyone's trying to solve in many, many ways.
00:11:44.320 --> 00:11:56.639
Right now, most of the solutions, they are the reason they're successful locally is because developers are able to do a local OAuth, run an MCP or a plug-in, and provide their own credentials, right?
00:11:56.799 --> 00:12:12.559
And which means that if you have like a company of 100 engineers, roughly like, I don't know, 90% are probably not even connecting the right tools and not seeing the the fullest extent the agents are how successful they are in accessing Versell or GCP and other tools, right?
00:12:12.799 --> 00:12:19.120
And that's where the enterprises are still like lagging behind in these capabilities because you don't have the infrastructure.
00:12:19.200 --> 00:12:21.840
Um, for example, you need sandboxes, and that's not enough.
00:12:21.919 --> 00:12:24.480
You don't want the agents to see those API keys.
00:12:24.559 --> 00:12:29.360
So you're gonna add some sort of like a networking proxy layer, like credentials proxy, right?
00:12:29.440 --> 00:12:34.639
So people are trying to solve that, like what the agent stack look like, the agent cloud for the background agents.
00:12:34.720 --> 00:12:38.480
And then the other question is how do you convince IT to bring differential access?
00:12:38.720 --> 00:12:48.399
Like who, when the when, like depending on which environment the agent is running, or for instance, which Slack channel, like what kind of credentials you want to bring to the agent, right?
00:12:48.559 --> 00:12:50.480
Um, so all those problems are still unsolved.
00:12:50.559 --> 00:12:56.879
Like we're very early, as I said, the models are way there, but the products that are really using these capabilities are still lagging behind.
00:12:56.960 --> 00:12:58.879
So a lot of opportunity right now.
00:12:59.440 --> 00:13:06.320
Yeah, and it's one thing if like you're like some people listening to this, they're like individual creators.
00:13:06.399 --> 00:13:08.240
Um, and then you can kind of move fast.
00:13:08.399 --> 00:13:15.039
But once you start to become a part of a larger organization, like the complexity goes up exponentially right now.
00:13:15.840 --> 00:13:16.320
That's right.
00:13:16.480 --> 00:13:22.879
The individuals are where like we're seeing the biggest, like the AI-pilled audience, like top commuters and all.
00:13:23.200 --> 00:13:26.480
And the enterprises feel like constrained on a couple of things.
00:13:26.639 --> 00:13:31.360
One is like the token budgets, like because most of these tools are successful locally.
00:13:31.600 --> 00:13:33.600
One of the things they don't have is auditability.
00:13:33.679 --> 00:13:38.559
Are these tokens being spent on their own projects or are people doing their own side projects?
00:13:38.639 --> 00:13:39.759
You don't have any idea, right?
00:13:40.000 --> 00:13:40.639
Oh, interesting.
00:13:40.799 --> 00:13:41.600
I didn't realize that.
00:13:41.679 --> 00:13:41.919
Okay.
00:13:42.080 --> 00:13:42.799
Like, yeah, yeah.
00:13:42.879 --> 00:13:44.720
I mean, for a startup like us, it's okay.
00:13:44.799 --> 00:13:49.519
Like, I mean, but for larger companies, they have to think about if some developers are spending like 5,000 bucks a month.
00:13:49.919 --> 00:13:51.279
Right, what are they spending it on, right?
00:13:51.440 --> 00:13:57.200
Are we even seeing the impact, or are they just doing some side projects and because it's so easy to do to do build anything now, right?
00:13:57.279 --> 00:14:00.720
You could be building your home home automation system for God's sake, like anything.
00:14:01.120 --> 00:14:02.960
Totally, which which we've all done, right?
00:14:04.240 --> 00:14:07.360
So, yeah, so so that's that's the biggest thing right now.
00:14:07.440 --> 00:14:15.840
Like a lot of the success has been like with the prosumer audience, like a lot of individuals and running ahead with like this hundred, two hundred order plans.
00:14:16.000 --> 00:14:25.360
Um, but the enterprise is still like the problem has not been fully solved in terms of the cost visibility and how do I operationalize the agents in terms of security.
00:14:25.679 --> 00:14:27.039
So a lot of gaps are still there.
00:14:27.120 --> 00:14:29.360
That's why the background agents have been a hit and a miss.
00:14:29.440 --> 00:14:40.399
People have been trying those ideas for the last uh more than a year now, but we haven't seen like big success in moving the token spent uh to cloud and the background agents.
00:14:40.720 --> 00:14:47.919
Okay, there's like code review and then there's like security review, security vulnerabilities.
00:14:48.240 --> 00:14:49.679
How related are they?
00:14:50.480 --> 00:14:53.120
I would say that two sides of the same coin.
00:14:53.360 --> 00:15:00.960
Like during code review, you're looking at um your best practices, your whatever the rails you have, and not and and of course security as well.
00:15:01.120 --> 00:15:10.159
You want to understand whether you are going to have PII leakage because these agents do make um a lot of mistakes that do compromise the security posture, right?
00:15:10.399 --> 00:15:24.960
When we look at security review, like you have had these like tools of the past um around um SCA and SBOM tools, a lot of like supply chain analysis, a lot of like dynamic tools as well, static analyzer like SemGrep and all.
00:15:25.120 --> 00:15:28.480
So those are getting supplanted with now a lot of AI-based analysis.
00:15:28.559 --> 00:15:36.399
So you could really reduce the false positives because that's been one of the biggest complaints with the SNCC and SAMGREPs of the world is like a lot of false positives.
00:15:36.480 --> 00:15:43.039
So people have, I don't know, thousands of findings in a mature repository, but they don't know what to prioritize or which ones are even real threats.
00:15:43.360 --> 00:15:50.399
Now with AI, some of that has been accelerated in doing a security review, and it's also like all almost like an arms race, which is very interesting.
00:15:50.480 --> 00:15:56.320
Now you're seeing really capable models are able to also find vulnerabilities, right?
00:15:56.480 --> 00:16:05.279
That's why the government's trying to regulate some of these like Mythos class and the bigger models because they're so good in finding zero-day problems.
00:16:05.440 --> 00:16:07.360
At the same time, your review has to catch up.
00:16:07.440 --> 00:16:08.480
It's like an arms race.
00:16:08.639 --> 00:16:14.799
I mean, you you will have to fight fire with fire, you have to now do security review with another AI.
00:16:14.960 --> 00:16:16.399
Um, so it's pretty exciting time.
00:16:16.480 --> 00:16:21.039
I think reviews and security reviews are really pretty much both sides of the same coin.
00:16:21.120 --> 00:16:23.039
It's really about the latency at the at this point.
00:16:23.200 --> 00:16:32.320
Like, yes, you want like a fast review on a PR that can finish in 10 minutes, but if it's an important feature, maybe you want to run a thorough security review that can take half an hour.
00:16:32.559 --> 00:16:40.399
It's just about spending time and more tokens in making sure that they're like uh no threats uh that can get in.
00:16:40.720 --> 00:16:47.120
And there's many different types of reviews because like you could even have like an agent like use, like if they're using the product, right?
00:16:47.200 --> 00:16:55.200
And trying to do prompt injection or trying to do other types of things to really try to jailbreak the product as well.
00:16:55.919 --> 00:16:56.799
Yeah, that's right.
00:16:56.960 --> 00:17:04.480
I mean, the if you're building an AA product, um securing them is also like challenging given how non-deterministic those systems are.
00:17:04.559 --> 00:17:12.319
So you could casual an agent, talk to it nicely, and it will give you spill all its secrets, including I don't know, it's system prompt or whatever credentials you put in.
00:17:12.400 --> 00:17:15.359
That's why it's not safe to add your credentials.
00:17:15.440 --> 00:17:18.799
And people have had so much problem with open claw, for example, right?
00:17:18.880 --> 00:17:26.400
I mean, it's great as in hobbyist, but if you're like really a serious enterprise, you have to take those concerns very seriously around guardrails.
00:17:26.559 --> 00:17:29.119
And um, I think we're like in very early.
00:17:29.200 --> 00:17:37.759
I mean, a lot of the products of the future are going to focus on how do we even like pick up the agents' reasoning traces and do analysis on them.
00:17:37.839 --> 00:17:42.640
Like, did these agents did they did these sessions got compromised in some way?
00:17:42.880 --> 00:17:50.720
So there's going to be like a lot of new startups um uh which we'll be tackling um say especially the agentic security.
00:17:51.200 --> 00:17:57.680
It's not about like a single prompt anymore, it's about the trajectory now, like even the evals and security looks very different now in the new world.
00:17:58.319 --> 00:18:05.599
What can a human do in code review that we cannot do with an agent yet?
00:18:06.559 --> 00:18:06.880
Right.
00:18:07.039 --> 00:18:08.799
So it's very interesting, right?
00:18:08.960 --> 00:18:15.279
I mean, if you look at the code reviews, um, they have certainly evolved than how they were two years back.
00:18:15.359 --> 00:18:21.519
So the type of mistakes an AI agent makes is not the same type of mistake a human makes.
00:18:22.079 --> 00:18:22.319
Right?
00:18:22.400 --> 00:18:28.559
So you're not looking at issues which are simple try-catch misses or some error handling is not there.
00:18:28.640 --> 00:18:31.440
Like the code is better than what most humans will write.
00:18:32.079 --> 00:18:32.319
Right?
00:18:32.559 --> 00:18:42.319
If you look at that code in isolation, now if you zoom out and look at the bigger picture, I mean, is this pull request something that is should be merged or not merged?
00:18:42.400 --> 00:18:45.119
Which is like, is it adding to the entropy of the code base?
00:18:45.200 --> 00:18:46.799
Is it reusing the right patterns?
00:18:46.880 --> 00:18:53.519
Is it um because in isolation it looks perfect, but half the time, I mean, you will see that it's like leaving some shadow paths behind.
00:18:53.759 --> 00:18:56.160
It's like changing the core flows of your product and so on.
00:18:56.559 --> 00:19:03.039
This chain that you're adding could make things slower or could add, you know, your memory leaks and all these other issues that could happen, right?
00:19:03.359 --> 00:19:03.680
Right.
00:19:03.839 --> 00:19:10.559
Because when agents are generating code, they have a very similar focus, taking the user's prompt and going all the way.
00:19:10.720 --> 00:19:17.279
And depending on how assertively you prompt the agent, they will just go out and do that thing, even whether it's the right thing or not.
00:19:17.519 --> 00:19:33.599
Now, as an enterprise, as a company, like I mean, you have a team of developers, some of them are good with prompting, so they don't like push the agents too much, but sometimes you will like force the agents to do something so unnatural that you have to detect that in the adversarial review against whatever your hard invariants are and so on.
00:19:33.680 --> 00:19:35.039
Are they getting violated, right?
00:19:35.279 --> 00:19:45.519
So the nature of review is definitely of human review, is moving away from line by line of code reviews because that's where AI is already doing a better job than humans, to higher order reviews.
00:19:45.599 --> 00:19:52.319
So you're looking at higher abstractions in understanding the architecture, understanding the other sort of blast radius impact of this change and so on.
00:19:52.480 --> 00:19:54.799
Yeah, one thing is sure the review is not gonna go away.
00:19:54.880 --> 00:19:56.799
Like, I mean, your guard deals never go away.
00:19:56.960 --> 00:19:58.720
For example, like Datadog is one example.
00:19:58.880 --> 00:20:02.000
Your cloud platforms are becoming better and better with time.
00:20:02.079 --> 00:20:07.519
Your Kubernetes makes sure you don't crash, but doesn't mean you don't you switch off your monitoring, right?
00:20:07.759 --> 00:20:10.960
So the so the review is gonna evolve, but it's never gonna go away.
00:20:11.839 --> 00:20:17.119
Now, Cursor bought a code review company, but for whatever reason that hasn't like really taken off.
00:20:17.279 --> 00:20:20.400
Like, why why why has that not taken off?
00:20:21.359 --> 00:20:22.720
That's a great question.
00:20:22.960 --> 00:20:28.640
So, so yeah, I mean they exited because also like because Code Rabbit was like still like a better product in many ways.
00:20:28.720 --> 00:20:39.680
And also that company started pre-gen AI and with a different philosophy that if you look at the time before pre-gen AI, humans were still reviewing, but the review was still a bottleneck.
00:20:39.839 --> 00:20:41.680
It's not like it wasn't a bottleneck.
00:20:41.920 --> 00:20:44.079
Yeah, it's been a bottleneck my whole life.
00:20:44.319 --> 00:20:44.960
Exactly.
00:20:45.200 --> 00:20:54.880
Sometimes you're waiting and you're like, I wanna I wanna push this, and you're like, you know, you've been you're waiting for your colleague to take a look at it, but your colleague's on vacation or something, or you know, whatever it might be.
00:20:55.119 --> 00:20:55.519
Right.
00:20:55.759 --> 00:20:59.759
So for that company, like the focus has been around graphite, right?
00:20:59.839 --> 00:21:01.759
So the focus had been around stacked pull requests.
00:21:01.839 --> 00:21:09.680
So the idea is like if the humans were in the review path, they can only review if the changes are small and byte-sized.
00:21:09.839 --> 00:21:20.160
So the idea is like you checkpoint your big feature into smaller pull requests or stack of commits that you can review one at a time and then stream all of them together, right?
00:21:20.319 --> 00:21:27.920
And the then the onus is on the author of a pull request to actually checkpoint the commits that way, to create a stack, which is meaningful enough for the reviewer to understand.
00:21:28.079 --> 00:21:31.519
For instance, the first commit of a stack is you made a schema change.
00:21:31.680 --> 00:21:36.480
The second commit could be you made um some business logic around it, then you add a API route.
00:21:36.559 --> 00:21:39.200
So you can actually stack your changes as you're building software.
00:21:39.359 --> 00:21:42.000
But that was like the idea, like pre gen AI, right?
00:21:42.079 --> 00:21:47.039
Now, with AI, you can't really you can ask the AI to stack, but usually it makes no sense.
00:21:47.119 --> 00:21:53.119
Like you can just ask it to go build a feature and let the reviewer figure out the right way of reviewing these changes.
00:21:53.279 --> 00:21:59.599
So that's where the disruption happened, like the techniques and the tools that we had designed around managing code changes.
00:22:00.079 --> 00:22:02.880
Before Gen AI looked very different with Gen AI.
00:22:03.279 --> 00:22:16.079
You think like maybe there's a way of like integrating the code review into your wherever you're storing your repo, GitHub or whatever it might be.
00:22:16.319 --> 00:22:25.519
Like you would think like somehow like the closer the integration there would be better, but for whatever reason, like GitHub doesn't really have a good solution either.
00:22:25.680 --> 00:22:30.400
Like, how do how do you see like the world evolving when it comes to code review?
00:22:30.799 --> 00:22:31.119
Right.
00:22:31.279 --> 00:22:35.279
I mean, this is a space which is a big addressable market.
00:22:35.440 --> 00:22:41.279
So platforms like GitHub will always compete with every other player in the market as well.
00:22:41.359 --> 00:22:43.920
I mean, it's almost like how Amazon was AWS, right?
00:22:44.079 --> 00:22:47.519
They had Cloudwash, but then still Datadog became very successful.
00:22:47.680 --> 00:22:55.200
So the way we say this market is there's going to be best of breed products like Code Rabbit, which will because a lot of companies care about best of breed.
00:22:55.279 --> 00:23:01.359
Like if there's some CFOs and say, okay, we anyways have a big contract with Microsoft, let's just buy everything from one place.
00:23:01.519 --> 00:23:07.920
But there are a lot of companies that go and buy best of breed solutions, like they'll buy Zoom, they'll buy Slack, right?
00:23:08.160 --> 00:23:10.079
And they're buying Code Rabbit, right?
00:23:10.160 --> 00:23:10.400
Yep.
00:23:10.640 --> 00:23:12.640
Um, so so that that's what we're seeing.
00:23:12.720 --> 00:23:24.160
I think this is one of those times where uh a lot of the VCs are also a little bit uh anxious about model labs going and winning everything, where we see kind of the world in a different way.
00:23:24.319 --> 00:23:28.319
We see there's a huge opportunity to solve problems with custom harnesses.
00:23:28.480 --> 00:23:38.319
Like code review is one such use case where Code Rabbit has built a custom harness which is like gives a best price to performance right now than anyone else in the industry.
00:23:38.559 --> 00:23:39.920
And that resonates very well.
00:23:40.079 --> 00:23:47.119
And that's why our like growth has been uh pretty much unobstructed despite GitHub competing with us over the last two years and so on.
00:23:47.680 --> 00:23:57.680
You have a very interesting lens because you're seeing code across lots of organizations and helping fix code across lots of organizations.
00:23:58.480 --> 00:24:10.079
Is there some kind of like in some ways like a co-op that you can that could happen where okay, if you can fix one in one organization, that can help you fix another organization?
00:24:10.319 --> 00:24:12.640
Like, does that somehow make your model better?
00:24:12.880 --> 00:24:18.799
Or is it just so proprietary that you can't actually use that data to the next one?
00:24:19.119 --> 00:24:22.640
Yeah, some of the learnings across orgs do carry forward.
00:24:22.720 --> 00:24:29.920
For instance, well, the reason why we have been further ahead of everyone is because we have a large install base in open source.
00:24:30.400 --> 00:24:35.039
And we every few seconds there's a pull request getting reviewed in open source.
00:24:35.200 --> 00:24:41.039
In fact, yesterday I was looking at like Kubernetes community sent out a newsletter saying that they love Code Rabbit, right?
00:24:41.119 --> 00:24:49.759
So we are in like very popular mission critical repositories right now, and that's like actually have uh help um acting as like a feedback loop.
00:24:50.079 --> 00:24:54.960
Yeah, that's great because you can use that, you're allowed to use it, and you don't have to, it's not proprietary.
00:24:55.119 --> 00:24:55.599
Yeah, okay.
00:24:55.920 --> 00:24:56.240
That's right.
00:24:56.400 --> 00:25:04.799
We are in a very unique position compared to a lot of the AA products out there which work in a more private setting, so you can't really train on their data in our case.
00:25:05.039 --> 00:25:18.480
We have been taking all the learnings we have from those runs, which are all in the sandbox, there is agent trajectories involved, and taking it back into improving the hardness, improving the context assembly, and even like choosing the right models that we want to use.
00:25:18.640 --> 00:25:22.079
Like we have so much data coming in that we are able to pick the right cocktail.
00:25:22.240 --> 00:25:26.559
Um, so we're not a product where we let the user pick a model from a drop-down.
00:25:26.640 --> 00:25:31.759
We we come up with the right cocktail of agent and subagents and different other models, right?
00:25:32.079 --> 00:25:33.440
And that's proved very valuable.
00:25:33.519 --> 00:25:37.200
So we're taking the learnings across these projects and helping everyone else out there.
00:25:37.279 --> 00:25:42.319
At the same time, the product also builds proprietary learnings within an organization.
00:25:42.400 --> 00:25:50.640
So there's still like a lot of tribal knowledge that does not translate well across organizations because sometimes people do things which are counterintuitive.
00:25:50.720 --> 00:25:55.519
I mean, the AI agent will say, okay, this is the best practice, but then someone will go and correct Code Rabbit.
00:25:55.680 --> 00:25:57.119
No, this is not how we do things.
00:25:57.359 --> 00:26:00.160
We are not your average data set, we are different, right?
00:26:00.319 --> 00:26:07.119
And Code Rabbit will remember that fact and then fine-tune the reviews based on those learnings for that organization.
00:26:07.680 --> 00:26:19.599
Okay, so when when we met you roughly two years ago and when you did your seed, okay, it was clear like you had a cool product, you guys were like super technical, you're like really into the you're into the product.
00:26:21.039 --> 00:26:28.400
What we didn't know then is also you guys are like masters of go to market and marketing um as well.
00:26:28.640 --> 00:26:34.400
Like, walk me through like the marketing playbook that you guys have used over the last couple of years.
00:26:34.799 --> 00:26:36.480
This is my third startup, first of all.
00:26:36.559 --> 00:26:41.039
So and go to market across all my startups has looked very different.
00:26:41.279 --> 00:26:46.079
Like my first startup was mostly in the cloud infrastructure space, so right when Kubernetes was taking off.
00:26:46.160 --> 00:26:48.799
So the GTM there was very content focused.
00:26:49.039 --> 00:26:54.480
Like getting into hacker news led to and led to a lot of like signups and meetings.
00:26:54.720 --> 00:26:57.039
We were mostly SDR focused in my first startup.
00:26:57.519 --> 00:27:03.119
This one has been all uh word-of-mouth, product-led growth company, right?
00:27:03.519 --> 00:27:18.960
And we kind of ground up, built that intentionally because this is one of those markets where we had seen previous successes where Cynic, for instance, is one company we took a leaf out of their playbook a little bit by making the product free for open source.
00:27:19.039 --> 00:27:24.240
And there was another reason when we started back in 2023, those were very early days of AI.
00:27:24.880 --> 00:27:28.799
And a lot of people didn't even have this new habit of using AI.
00:27:29.279 --> 00:27:29.599
Right?
00:27:29.680 --> 00:27:36.720
So we wanted to make sure that uh light as many fires as we can, like get the product out there so that people build this new habit.
00:27:37.039 --> 00:27:37.279
Right?
00:27:37.359 --> 00:27:40.640
That was a biggest battleground more than pricing and other things.
00:27:40.720 --> 00:27:46.799
It was like, would people accept or reject AI in in pull request workflow, which is very serious, right?
00:27:47.279 --> 00:27:54.640
Um, so one of the strategies we did there was to open up the product for open source for a variety of reasons.
00:27:54.799 --> 00:27:59.839
One is wanted to show people that how effective this product is, like seeing is believing.
00:27:59.920 --> 00:28:03.440
So those kinds of products, you can't just put a message on website and make promises.
00:28:03.680 --> 00:28:05.200
You have to really show people, right?
00:28:05.359 --> 00:28:07.839
And open source was the best showcase we could find.
00:28:08.160 --> 00:28:10.960
Second is open source maintainers actually had a pain point.
00:28:11.039 --> 00:28:15.920
I mean, there were a lot of like contributions from unknown people, and this was like a thankless job.
00:28:16.000 --> 00:28:18.880
So they were already struggling with code reviews.
00:28:19.039 --> 00:28:24.960
And thirdly, we knew that the open source adoption is going to set the flywheel for constant product improvement.
00:28:25.119 --> 00:28:27.680
So we kind of like hit two or three birds with one stone.
00:28:28.000 --> 00:28:28.240
Right.
00:28:28.319 --> 00:28:32.640
You're making your marketing and you're making your product better on one big thing.
00:28:32.880 --> 00:28:33.279
That's right.
00:28:33.359 --> 00:28:35.759
And these open source projects were like almost like influencers.
00:28:35.839 --> 00:28:39.759
These maintainers are influencers, they have a lot of weight in on Twitter and other places.
00:28:39.839 --> 00:28:45.200
And so one door led to another, and we got like viral adoption in open source.
00:28:45.359 --> 00:28:48.960
We also had a lot of viral adoption in the early days in Japan, very interestingly.
00:28:49.039 --> 00:28:54.799
So we had no boots on ground, but we had I don't know 30 blog posts written by end users about how good the solution is.
00:28:55.039 --> 00:28:56.160
Yeah, we noticed that too.
00:28:56.240 --> 00:28:59.200
Like, why why did the Japan thing take off the way it did?
00:28:59.599 --> 00:29:00.960
That we will never find out.
00:29:01.039 --> 00:29:05.440
Like, I mean, we don't it just got viral independently.
00:29:05.519 --> 00:29:12.079
We had no hand in that, it just went viral on its own, while it also went viral in the US, thanks to open source.
00:29:12.319 --> 00:29:15.839
And it was also like I think the time in the market, like we were early movers.
00:29:15.920 --> 00:29:19.759
Uh, people, it was AI was a novelty, so a lot of people were curious.
00:29:20.000 --> 00:29:22.240
At the same time, the product lived up to its promise.
00:29:22.319 --> 00:29:24.799
It wasn't just slop, it was actually adding real value.
00:29:24.960 --> 00:29:29.440
So a lot of things came together, and even on the marketing, we were like very influencer focused.
00:29:29.599 --> 00:29:34.079
So we would work with top content creators to talk about Code Rabbit.
00:29:34.160 --> 00:29:37.440
We had like, for example, fire ship video um placements and all.
00:29:37.519 --> 00:29:40.079
So that always led to like signups and awareness.
00:29:40.160 --> 00:29:44.640
Uh we have billboards, like you have seen SF, like we have like we did the station takeovers.
00:29:44.799 --> 00:29:46.640
Um, so all those things helped a lot.
00:29:46.880 --> 00:29:56.400
I was in uh the Amtrak station in New York, taking a train to DC, and the whole station was was wrapped in code rabbit.
00:29:56.480 --> 00:30:00.559
And I'm like, wait, why, why it like are there even that many developers?
00:30:00.640 --> 00:30:03.440
Like, what was your reasoning to do something like that?
00:30:03.759 --> 00:30:05.039
Yeah, it's about the mind share.
00:30:05.119 --> 00:30:09.200
I think we just wanted to be in the series where we think there's like developer count um population.
00:30:09.359 --> 00:30:10.559
We did one in Austin, I believe.
00:30:10.640 --> 00:30:14.000
We did in Bangalore, we are running ads, we're running in in the Bay Area, of course.
00:30:14.319 --> 00:30:15.920
Um, South Bay SF.
00:30:16.319 --> 00:30:19.039
Um, and we know that the pain point is so prevalent.
00:30:19.119 --> 00:30:20.559
Like, this is a prosumer product.
00:30:20.640 --> 00:30:24.240
So if you look at our customer base, it's millions of developers that we want to reach out to.
00:30:24.400 --> 00:30:25.440
Yeah, you even have a free price.
00:30:25.680 --> 00:30:34.640
It starts, you know, you can have a freemium version, and then people like upgrade to uh a little bit more, and then you can upgrade to a little bit more, and then all of a sudden, like the whole organization starts using it.
00:30:34.720 --> 00:30:36.559
Is that that's kind of typically the way it works?
00:30:36.799 --> 00:30:48.319
Like, like one developer starts using it, they tell another one inside, they tell a few more, then you've got a pod of 10 and then kind of spreads, and then eventually you do like an enterprise deal.
00:30:48.720 --> 00:30:49.200
That's right.
00:30:49.279 --> 00:30:52.559
It's it's always about getting people hooked into this new habit.
00:30:52.720 --> 00:31:03.839
So a lot of times we get pulled into these larger deals because some developers in that company have tried Code Rabbit in their own personal repositories or some open source project that they care about or they contribute to.
00:31:04.000 --> 00:31:08.240
And that's usually like how we get into like we don't have a very strong outbound operation.
00:31:08.319 --> 00:31:09.599
It's all inbound for us.
00:31:09.759 --> 00:31:11.279
And that's that's a primary driver.
00:31:11.359 --> 00:31:19.039
If when we do a survey on where did you hear about us, majority would be through the developers in that company who tried out in open source or something.
00:31:19.279 --> 00:31:21.359
And so there's a very effective GTM motion.
00:31:21.519 --> 00:31:26.240
I don't know how much it translates to other domains, but for DevTooling, this has worked really, really well.
00:31:27.039 --> 00:31:36.400
You have like very, very big companies that are already adopting um, like some of the more iconic companies are already adopting Code Rabbit.
00:31:37.200 --> 00:31:50.079
Uh, it's interesting to go from like this prosumer to a very enterprisey thing where I'm sure you need lots of systems and you need to review things and admin privileges.
00:31:50.240 --> 00:31:56.640
And uh, like how does one decide how to build the products for this like wider enterprise?
00:31:57.119 --> 00:31:58.880
It's a very interesting challenge.
00:31:59.039 --> 00:32:08.240
And you know, the product does get complicated as you add more knobs to it, which at the same time um become too overwhelming for the prosumer audience.
00:32:08.400 --> 00:32:13.440
Yeah, they like simplicity and striking that right balance is always been like a big debate.
00:32:13.599 --> 00:32:16.160
I mean, of course, you have to build the go-to market motion first, right?
00:32:16.240 --> 00:32:22.799
I mean, so for the longest time our revenue had been mostly PLG self-serve revenue.
00:32:23.039 --> 00:32:35.279
And since then, we have started building a very successful sales engineering organization, um, a lot of like enterprise AAEs and further segmenting them, like mid-mark, mid-enterprise, strategic, like we've done all the segmentation, right?
00:32:35.599 --> 00:32:37.759
Um, then the other thing is a product, right?
00:32:37.839 --> 00:32:39.839
The product has to be enterprise ready.
00:32:40.079 --> 00:32:48.240
And we have customers who want self-hosted and they want to uh be uh want us to be in Europe, like whatever the GDPR reasons, right?
00:32:48.400 --> 00:32:49.839
And so we you have to do all that.
00:32:49.920 --> 00:32:55.119
Like now with the agents, it's also so easy to build all those configurability that people need.
00:32:55.440 --> 00:32:56.960
The real trick is in the product layer.
00:32:57.039 --> 00:33:08.640
Like, how do you keep the product simple, but also like still programmable enough that the enterprises, even on the legacy code basis with different um velocities of shipping, can still tailor it, right?
00:33:08.880 --> 00:33:10.799
So, yeah, I mean, everything has to go hand in hand.
00:33:11.119 --> 00:33:17.759
I mean, you still going like essentially from zero to 50 plus million in two years is it's kind of insane, though, still, right?
00:33:17.839 --> 00:33:22.160
Like, and like even just an organization, you're hiring all these people.
00:33:22.400 --> 00:33:32.160
Like, how do you yeah, maybe you're not hiring as many people as like a company would have done 10 years ago if they're going that fast, but you still have a lot of people you have to absorb and train.
00:33:32.319 --> 00:33:37.599
Like, how do you think about just just generally growing a company in that kind of hypergrowth mode?
00:33:37.920 --> 00:33:42.319
No, that's a good like we have 250 people, so it's not like the AI has eliminated.
00:33:43.599 --> 00:33:51.519
That's a lot of people, and you gotta train them, you gotta motivate them, you have to get them excited, you have to get them all on the same page, right?
00:33:51.759 --> 00:33:55.119
Uh just even like interviewing that many people, a lot of people.
00:33:55.519 --> 00:34:07.359
Yeah, that's why you have to scale with like uh professional leadership, like and the needs change as you went from let's say zero to one million, one to ten was a different need, and then 10 to 50 has been a different need.
00:34:07.680 --> 00:34:14.480
Like, so some of those needers uh either they up-level themselves or we have to bring in um other leaders on top, right?
00:34:14.960 --> 00:34:19.199
And tap into their network to hire more A players and so on.
00:34:19.440 --> 00:34:22.400
And that's what we have done with the sales team and product teams and it's run.
00:34:22.639 --> 00:34:24.960
Like, but on the engineering side, I would say we have not grown that much.
00:34:25.039 --> 00:34:32.480
Now it's very interesting that we are living in a time where our shipping velocity has exceeded our ability to market those products.
00:34:32.559 --> 00:34:33.679
This is the first time I'm seeing that.
00:34:33.840 --> 00:34:40.400
Usually you will have marketing run ahead, promise the features which are like a quarter or two away, and it's flipped.
00:34:40.480 --> 00:34:45.679
Now we have already shipped the changes, and marketing is strike trying to figure out what happened, right?
00:34:46.159 --> 00:34:47.039
It's funny, yeah.
00:34:47.199 --> 00:34:52.320
You you're usually selling like three quarters ahead, and now you're selling like one or two quarters behind.
00:34:52.639 --> 00:34:54.079
That's the reality, very interesting.
00:34:54.159 --> 00:34:59.760
And not just us, even the model labs, they're shipping, building in public, broadcasting live.
00:34:59.840 --> 00:35:03.199
So the marketing nature of marketing has changed, right?
00:35:03.360 --> 00:35:08.320
And you don't need like a big engineering organization to keep your go-to market team busy.
00:35:08.400 --> 00:35:11.599
You need a lot of sales because you still need to have those meetings.
00:35:11.679 --> 00:35:19.920
You can't automate with AI, but engineering and marketing to some extent, content, we are seeing rapid um shipping velocity thanks to AI.
00:35:20.079 --> 00:35:22.639
Um, but on the ground, you still need sellers.
00:35:22.719 --> 00:35:24.079
You do you need that DNA.
00:35:24.400 --> 00:35:35.440
You have to still sell the value, the ROI, the why to the end user and customers, especially the leadership in the enterprise teams, where the more stakeholders involved and get those deals to the finish line, right?
00:35:35.599 --> 00:35:38.320
The POC and the bottom swell is never enough.
00:35:38.559 --> 00:35:42.480
Like you have to find the economic buyer and convince them as well, right?
00:35:43.119 --> 00:35:51.679
Now there's there's this consensus, or some people believe AI is gonna wipe out or or significantly reduce the junior engineers.
00:35:51.760 --> 00:35:59.119
And I've heard you say it's kind of closer to the opposite that new grads can do now in two hours, but used to take two weeks.
00:35:59.280 --> 00:36:03.599
Like, what does the entry-level job look like in a few years?
00:36:03.920 --> 00:36:10.320
We you still need engineers of all kinds of calibers, like the senior engineers definitely bring in experience and system level knowledge.
00:36:10.400 --> 00:36:14.079
And it's about like the pod that you want to create when you're delivering software.
00:36:14.159 --> 00:36:28.239
So we have started like organizing our teams into smaller um these units called pods, where you have someone in the driving seat with an agent, but you have the UX team in the same room, you have the product team, they're all touching the code, by the way, but their own PCs, right?
00:36:28.480 --> 00:36:32.719
So the interns, we also have a lot of interns and we are able to make them successful.
00:36:32.880 --> 00:36:42.960
In fact, what I love about interns is they are just limited by imagination at this time, like whereas some of the engineers will still try to go a little bit deeper into the code, not willing to let go.
00:36:43.199 --> 00:36:50.159
Try to like do more tactical driving when you could now take a step back and let the agent also make a few decisions and run faster.
00:36:50.320 --> 00:36:56.480
So we're seeing that imbalance where sometimes even the senior people are being held back because they're too afraid to let go.
00:36:56.719 --> 00:36:57.039
Yeah.
00:36:57.280 --> 00:36:57.519
Yeah.
00:36:57.599 --> 00:36:58.000
Right.
00:36:58.239 --> 00:37:06.639
And at the same time, you also need their experience to watch the backs on what to review and where to how to validate, how to scale, because it's easy to prototype.
00:37:06.800 --> 00:37:10.400
But some of these features, once they go out, you still have to solve the scaling problems, right?
00:37:10.480 --> 00:37:12.400
Which you only discover in production.
00:37:12.719 --> 00:37:12.960
Yeah.
00:37:13.039 --> 00:37:17.039
So you need like people who are great with imagination and curiosity.
00:37:17.199 --> 00:37:24.000
You need engineers who understand distributed system scalability, you need product UX, everyone to come together and touch code at this point.
00:37:24.079 --> 00:37:27.519
So everyone is coding in a way, just their own PCs.
00:37:27.840 --> 00:37:42.400
Now, when I started my career writing in Java, then maybe in Scala or something, um, I I always read code, but I never in my life have ever read assembly, or certainly never read machine level code.
00:37:42.719 --> 00:37:49.920
Are we going to move to a world where nobody reads code just like when I started a career, no one read assembly?
00:37:50.159 --> 00:37:52.079
I think we are already almost there.
00:37:52.159 --> 00:37:58.000
Like, even I am mostly using my agent or code rabbit to ask the right questions about the code.
00:37:58.239 --> 00:38:05.199
Because as an architect level, like I usually just with a few questions, I'll have an idea that this is going in the right direction or not.
00:38:05.440 --> 00:38:05.760
Right?
00:38:05.840 --> 00:38:08.880
It's just about talking to the agent now more than writing the code.
00:38:08.960 --> 00:38:10.400
Um, and that quality is already there.
00:38:10.480 --> 00:38:13.199
And a lot of this is going to be codified in tools like Code Rabbit.
00:38:13.280 --> 00:38:16.719
So you don't need an expert person to do all this in every company.
00:38:16.800 --> 00:38:18.239
So the tools can do it for you.
00:38:18.480 --> 00:38:24.159
But yeah, I mean, I think the that time is just around the corner when you are really not reading the code.
00:38:24.239 --> 00:38:26.559
It's like taking the steering wheel off the car.
00:38:26.639 --> 00:38:28.000
We are almost at that point.
00:38:28.159 --> 00:38:28.480
Yeah.
00:38:28.719 --> 00:38:33.039
And we have already seen that IDEs are not future-proof, right?
00:38:33.199 --> 00:38:36.719
I mean, I have the first time I have moved out of my terminal.
00:38:36.880 --> 00:38:39.840
I used to be a NeoVim person for the last 30 years.
00:38:40.000 --> 00:38:42.880
And I've not been in my terminal in the last six months.
00:38:42.960 --> 00:38:44.559
I'm in Codex Desktop app.
00:38:44.800 --> 00:38:45.280
Yeah, yeah.
00:38:45.519 --> 00:38:49.760
Yeah, which is also beautiful and easy to use and kind of fun, right?
00:38:50.000 --> 00:38:52.000
Yeah, and just talking to the agent, right?
00:38:52.079 --> 00:38:55.760
And you need tools like Code Rabbit, like to do an adversarial review.
00:38:55.840 --> 00:38:57.039
So I have a loop set up.
00:38:57.119 --> 00:39:01.119
So until Code Rabbit is happy, we just keep committing, right?
00:39:01.519 --> 00:39:07.760
And and and then once everything is great, and if there's a feature is not high risk, you don't need to involve another human.
00:39:07.840 --> 00:39:09.440
You can just ship it, right?
00:39:09.599 --> 00:39:16.800
So we started getting your confidence that even an agentic approval is good enough for 80% of your features.
00:39:17.280 --> 00:39:20.719
There's one that okay, there's a review, there's a code review.
00:39:20.960 --> 00:39:28.960
Um, there are a lot of other people are using whether it's codecs or callers to create content.
00:39:29.599 --> 00:39:39.519
Um, so I've got like I've got long-running codecs tasks, which are essentially creating content for me, um, doing lots of in some ways that's kind of like code.
00:39:39.760 --> 00:39:47.280
And I could use a review system to review my content, just like people review their code in a way.
00:39:47.440 --> 00:39:53.519
Like, could you imagine there's like a like analogy to review content?
00:39:54.320 --> 00:39:55.760
Yeah, it's a very interesting thing.
00:39:55.840 --> 00:40:01.519
Like any problem today that can be translated to a coding problem, you're gonna move 10x faster.
00:40:01.679 --> 00:40:04.079
So if you can run your marketing as code, yeah.
00:40:04.239 --> 00:40:06.559
I mean, the idea is like can you run your company as code, right?
00:40:06.719 --> 00:40:07.280
Yeah, exactly.
00:40:07.519 --> 00:40:17.599
And if you can eval against something, uh it can be your design taste or something, the agent can loop around and get to really a picture perfect one pager, slide deck, whatever it is, right?
00:40:18.000 --> 00:40:20.400
And we are also we've been moving in that direction.
00:40:20.480 --> 00:40:24.239
In fact, I have codec sessions running, doing sales deck, one pagers, right?
00:40:24.400 --> 00:40:26.320
In in Markdown and MARP.
00:40:26.480 --> 00:40:30.079
So I use this MARP as one of the tools where it creates slides, right?
00:40:30.320 --> 00:40:35.280
Um, so the idea there is like the humans still need to understand these changes.
00:40:35.360 --> 00:40:43.360
So if these agents are gonna run longer and longer periods of time, there's gonna be a new concept of explainability that's going to emerge.
00:40:43.519 --> 00:40:48.880
Just like how Datog did observability, there was a lot of chaos in fast-moving environments, just like that.
00:40:48.960 --> 00:40:50.320
There is going to be a lot of chaos.
00:40:50.800 --> 00:40:58.079
Agents are gonna create massive amounts of work, and you're gonna be bottlenecked on reviewing and understanding those outputs.
00:40:58.239 --> 00:40:58.480
Yeah.
00:40:58.719 --> 00:41:01.119
Like whether it's content or even information retrieval.
00:41:01.199 --> 00:41:05.679
You want to like build trust in the output, is like, did this agent even go and look at the right data?
00:41:05.920 --> 00:41:10.719
If you are asking it to get your I don't know, weekly website visitors, did it even access the right tool?
00:41:10.800 --> 00:41:12.719
Right, you want to build trust, right?
00:41:12.880 --> 00:41:27.280
And I think there's going to be, like, as I said, a lot of products that are gonna take agents, reasoning traces, or reverse engineers from the output, just like how we do in review, we kind of reverse engineer what could have happened to build sort of an explainability walkthrough.
00:41:27.519 --> 00:41:32.239
And that's what the humans are gonna review, whether you're doing content generation, because everything is code.
00:41:32.320 --> 00:41:35.039
Code is now the execution layer for these agents, right?
00:41:35.360 --> 00:41:41.039
And so you need reviews not just for your software, but also for content now, because everything is code end of the day.
00:41:41.360 --> 00:41:42.639
But that's exactly what I'm trying to.
00:41:42.719 --> 00:41:45.039
I I did a long-running codex task.
00:41:45.119 --> 00:41:48.480
It took about five days, it created all this content for me.
00:41:49.039 --> 00:41:52.639
But honestly, I in the end, like I didn't really understand what happened.
00:41:52.800 --> 00:41:54.639
I'm like, wait, what did you do?
00:41:54.719 --> 00:42:10.079
Like, and then I'm trying to like it was very, very hard for me to really get a handle on what it did and what was the level I could like go in and inspect some things, but it was it was like almost like a very it was almost like a code review, like a human code review that I had to go in and do.
00:42:10.400 --> 00:42:15.199
When will something like Code Rabbit be able to do that for me, or can it already do it?
00:42:15.360 --> 00:42:16.719
And I just don't understand it.
00:42:17.039 --> 00:42:18.960
I mean, with the agents, you could do it.
00:42:19.039 --> 00:42:23.840
Like the thing is that for us, our audience primarily has been the developer audience, right?
00:42:23.920 --> 00:42:30.239
And we have been like bringing some of the products into other surfaces given how much how many installs we have in GitHub.
00:42:30.320 --> 00:42:36.480
So it's kind of natural for us to go from GitHub into Slack and other systems like Jira Linear.
00:42:36.639 --> 00:42:41.199
Now, Jira Linear are losing importance, interestingly, but Slack is becoming even more important, is what we're saying.
00:42:41.440 --> 00:42:41.920
Yes, for sure.
00:42:42.000 --> 00:42:42.719
Yeah, right.
00:42:43.280 --> 00:42:51.440
And those are the surfaces where we are trying to bring general purpose agents, which is building on the code context, but then you can use them to other work.
00:42:51.679 --> 00:42:54.639
And one of the things we are doing is instrumenting those agents.
00:42:54.800 --> 00:43:08.320
So one of the cool things we have done is like because we control the agents, it's a background agent, we are picking up I know tens of megabytes of reasoning traces out of these runs and running layers of AI on top to review those, right?
00:43:08.400 --> 00:43:14.239
And we're building visualizations to understand not just your um coding outputs but knowledge work as well.
00:43:14.559 --> 00:43:22.320
So building a more generalized explainability layer that goes beyond code reviews into any kind of agentic run visualization.
00:43:22.559 --> 00:43:24.239
And we are very bullish about the future.
00:43:24.320 --> 00:43:31.360
I think the explainability is going to be even a bigger category than what observability was in the past.
00:43:31.599 --> 00:43:36.639
And it's gonna start with code review because that's been the first biggest PMF area software agents.
00:43:37.360 --> 00:43:40.800
But very soon people are gonna realize that they have to run marketing as code now.
00:43:41.119 --> 00:43:46.800
Accounting as code, legal as code, when you and we need to explain all these changes all the way through.
00:43:47.280 --> 00:43:47.440
Right.
00:43:47.679 --> 00:43:52.079
Just like how you did infrastructure as code, you're gonna start committing your marketing assets in GitHub.
00:43:52.480 --> 00:43:54.639
So it's all gonna end up in GitHub eventually, right?
00:43:54.800 --> 00:43:55.519
Yeah, yeah.
00:43:55.840 --> 00:43:57.440
And by the way, it'd be very helpful.
00:43:57.519 --> 00:44:10.079
I mean, like if you if you have a legal agreement, if Very helpful to have uh you know be able to go back in time and understand like your commits to your legal agreement and how those changed over time and who committed what.
00:44:10.159 --> 00:44:20.480
And in a legal agreement, you may be writing contract with your customer and you both might be committing to it, and you have to resolve these kind of commits in some sort of way as well.
00:44:21.119 --> 00:44:21.840
Yeah, that's right.
00:44:21.920 --> 00:44:28.480
The way we are going is like, I don't like a lot of people on Twitter, some thought leaders will say, okay, the PR is dead, like you don't need pull requests.
00:44:28.559 --> 00:44:29.280
But we disagree.
00:44:29.360 --> 00:44:35.840
I think it's going to become the center of gravity in this entire movement in many, many ways, because there's no choke point.
00:44:35.920 --> 00:44:41.039
You can't just ship product without taking a look first, and especially if you're in a team setting, right?
00:44:41.280 --> 00:44:47.519
So this central quality gate is going to try to go beyond even software reviews that we are doing right now to other things.
00:44:47.760 --> 00:44:51.760
And yeah, I mean, I think the review step is going to be the most important step.
00:44:51.840 --> 00:44:55.039
That's what humans are going to end up doing, just doing the work, right?
00:44:55.280 --> 00:44:58.719
And maybe upstream giving some taste and collaborating on the taste as well.
00:44:58.880 --> 00:45:07.440
So it's either like you're collaborating on what to build and reviewing the prompts that the agents will run, or you're going to end up reviewing the end output.
00:45:07.920 --> 00:45:19.280
It's interesting because so many people I know who have never coded before, just a you know, even a few months ago, they're now like they're now using things like they didn't even have a GitHub account.
00:45:19.360 --> 00:45:20.559
Now they have a GitHub account.
00:45:20.719 --> 00:45:27.039
They have all these kind of things that are so you can start to see these like other tools that are out there really explode.
00:45:27.199 --> 00:45:34.320
It seems like the Microsoft acquisition of GitHub, which was, I don't know, seven years ago-ish, seems to be a very good thing.
00:45:34.400 --> 00:45:40.239
Like there seems to be a lot of these like tools that are in that will be enduring going forward.
00:45:40.719 --> 00:45:41.519
Yeah, that's right.
00:45:41.679 --> 00:45:45.519
I mean, I think GitHub was a great pick by Microsoft a few years back.
00:45:45.599 --> 00:45:49.840
No one saw it coming that how important this is going to become to the agent economy, right?
00:45:49.920 --> 00:45:50.559
Going forward.
00:45:50.800 --> 00:45:53.920
And of course, people are trying to build serverless Git.
00:45:54.000 --> 00:45:57.199
I mean, Cursor is launching a competitor, so are like a few players.
00:45:57.280 --> 00:45:59.599
Um, but it's almost going to be like very hard.
00:45:59.679 --> 00:46:01.679
Like it's uh it's kind of the blue sky moment.
00:46:01.760 --> 00:46:06.559
It's once you have the ecosystem, the tooling around it, the open source is all in GitHub.
00:46:06.639 --> 00:46:09.119
So it's gonna be very hard for another platform to emerge.
00:46:09.360 --> 00:46:22.000
And you want your like your repo to be independent of your you like because if you move from cursor to something else, you don't want to all of a sudden like have to like you you want it to almost be an independent system.
00:46:22.559 --> 00:46:23.519
That's right, right.
00:46:23.679 --> 00:46:26.320
And you want your reviewers to be independent.
00:46:26.400 --> 00:46:29.440
Um lot of people buy Code Rabbit because of that reason.
00:46:29.679 --> 00:46:34.239
Is if you were to use your coding agents for review, their context is already polluted, they're like too agreeable.
00:46:34.320 --> 00:46:39.039
Like once you have a session, it wouldn't question its own judgment and decisions, right?
00:46:39.280 --> 00:46:45.519
So you need a very adversarial clean session, which is programmed in a different way with different guidelines.
00:46:45.760 --> 00:46:47.280
And also it can be very thorough.
00:46:47.440 --> 00:46:58.400
You don't want to do that inline on the developer laptop, like you want to do that in a pull request model in a background agent that can run for 15-20 minutes and be very thorough before some things go out.
00:46:58.719 --> 00:47:00.320
Now, a couple personal questions.
00:47:00.480 --> 00:47:05.840
Um, okay, the flex capital, we're we're very happy investors in uh Code Rabbit.
00:47:06.079 --> 00:47:10.960
We're also investors in Tracer, which is run by your brother Tanveer.
00:47:11.039 --> 00:47:13.119
Okay, what did your parents do well?
00:47:13.280 --> 00:47:15.199
Because obviously, like they did something well.
00:47:15.280 --> 00:47:16.880
So, what did they do well?
00:47:17.360 --> 00:47:19.199
How are you guys raised?
00:47:20.079 --> 00:47:22.239
No, we've just been like very entrepreneurial.
00:47:22.320 --> 00:47:27.599
Like we started the first company together that was back in 2014 called Nexil.
00:47:28.159 --> 00:47:32.159
Um, and then we did second company together called Flux Ninja, which by the way didn't do very well.
00:47:32.239 --> 00:47:35.440
We ended up um Code Rabbit kind of came out of that startup as well.
00:47:35.599 --> 00:47:35.920
Yep.
00:47:36.079 --> 00:47:38.639
Um, so yeah, so me and my brother have been like very close.
00:47:38.960 --> 00:47:42.960
Was there something your parents did that got you entrepreneurial?
00:47:43.360 --> 00:47:44.719
It's about risk taking, right?
00:47:44.800 --> 00:47:50.559
So the one of the things the parents can teach you is to either be like risk averse and maybe get a job somewhere.
00:47:50.719 --> 00:47:57.519
Like that's what most people say, go get an MBA and getting hired by maybe Google is seen like at like 10 years, 15 years back, everything.
00:47:57.599 --> 00:47:58.159
That was the thing.
00:47:58.239 --> 00:48:03.519
Like you were going into a great brand name company, starting your career, whether it's consulting or something, right?
00:48:03.679 --> 00:48:10.719
Um, but my parents were more like just they were like they always like very much um risk focused.
00:48:10.800 --> 00:48:12.960
Like, I mean, the higher the risk, the higher the reward.
00:48:13.119 --> 00:48:15.280
And that's the game we learned to play growing up.
00:48:15.440 --> 00:48:22.320
Like, whether it was the schooling that we got, even going to the schools that were more focused on leadership and those qualities, right?
00:48:22.400 --> 00:48:26.000
Uh, to even the culture at home, it was about what's the worst gonna happen?
00:48:26.079 --> 00:48:27.679
Like, you always have something to fall back to.
00:48:27.840 --> 00:48:28.000
Yeah.
00:48:28.159 --> 00:48:32.000
So we could be like pretty much unshackled and take risk early days in our career.
00:48:32.079 --> 00:48:35.519
And even now we take risk, it's just like different appetite.
00:48:35.679 --> 00:48:39.840
Um, so I think that would be the biggest thing that I would say we did growing up.
00:48:40.159 --> 00:48:46.320
Now, if you're gonna give advice to an 18-year-old for their career, you know, it's uncertain time.
00:48:46.400 --> 00:48:49.280
A lot of 18-year-olds are not sure what they should be doing.
00:48:49.360 --> 00:48:51.119
Like, what advice would you be giving?
00:48:51.440 --> 00:48:57.840
Right now, I would say is the best time I've seen in the history to do startups and create value, right?
00:48:58.239 --> 00:49:01.119
These inflection points are very, very rare.
00:49:01.280 --> 00:49:05.280
Like it's uh once in a decade or two decades kind of an opportunity.
00:49:05.840 --> 00:49:12.639
And there is just so much, like I would say, land grab opportunity in terms of the use cases.
00:49:12.880 --> 00:49:14.800
We are like in the very early innings of AI.
00:49:14.960 --> 00:49:20.639
Like what we are building, it's not even 1% of what these models even today are capable of, right?
00:49:20.960 --> 00:49:22.559
So you're limited by imagination.
00:49:22.639 --> 00:49:28.400
Like, right now is the time if you are like not financial, there's no financial reason to join like or pay off the loans.
00:49:28.639 --> 00:49:30.239
Uh go out and take a risk, right?
00:49:30.320 --> 00:49:40.159
And the lot of VCs in the Bay Area, like, I mean think this space, this place is very special in terms of even how much uh people are willing to bet on the young talent.
00:49:40.480 --> 00:49:44.960
Like if you look at the cursor founders, cognition, they're all like young talent, like on the younger side.
00:49:45.199 --> 00:49:56.239
Yeah, um, because even VCs have realized that they are the ones which are moving fast with AI compared to a lot of the experienced folks who are now still like hesitant hesitant on adopting and questioning the priors, right?
00:49:56.480 --> 00:50:02.400
So I think it's a great time for like 18-year-olds, like first of all, learn how to use these agents really effectively.
00:50:02.480 --> 00:50:06.639
Even a single person can now be effective as I don't know, 20, 30 people, right?
00:50:06.800 --> 00:50:09.519
Very easily, even if you're doing it in the right way.
00:50:10.639 --> 00:50:14.639
How are you personally using AI like in your personal life?
00:50:15.760 --> 00:50:25.519
So, in my personal life, I mean, beyond coding, I mean, I am trying to use codecs more than chat GPT, I would say, given that code is now the foundation of everything.
00:50:25.840 --> 00:50:28.000
So, so that's become my go-to tool.
00:50:28.079 --> 00:50:38.639
Like early, like nine months back, I used to go to chat GPT, use sometimes a pro model as well, like especially when I'm doing some deeper research into planning something uh in my house or something.
00:50:38.800 --> 00:50:41.519
Uh, but now it's becoming more and more code-oriented, right?
00:50:41.760 --> 00:50:49.840
Yeah, I found like even planning a vacation using codecs is better than planning a vacation using chat GPT.
00:50:50.559 --> 00:50:51.119
That's right.
00:50:51.199 --> 00:50:51.599
That's right.
00:50:51.679 --> 00:51:02.480
We're seeing like the coding agents are going to be maybe the main substrate now, the underlying platform powering uh all kind of work or non-work, um, coding or non-coding kind of use cases.
00:51:02.639 --> 00:51:04.480
Um, that I'm already seeing in my personal life.
00:51:04.559 --> 00:51:06.880
I mean, unfortunately, I don't get much time for personal stuff.
00:51:07.199 --> 00:51:07.679
Of course, yeah.
00:51:07.920 --> 00:51:15.519
My wife's been meaning to like having my own, like I've two daughters like to teach them coding at some point, especially with codecs, not even finding time.
00:51:15.679 --> 00:51:17.199
Uh, but overall, yeah.
00:51:17.280 --> 00:51:20.320
I mean, in terms of AI, I use Grok a lot in my Tesla.
00:51:20.400 --> 00:51:27.519
So I've got Grok set up, so it's so convenient, like on the wavy, um, just to keep the kids busy, we'll just talk to Grok.
00:51:28.000 --> 00:51:31.280
And and you just just for fun, ask questions or keep it going.
00:51:32.000 --> 00:51:35.519
Like, it's really good with grams, like you want to like do mad libs or something.
00:51:35.599 --> 00:51:36.000
It's fun.
00:51:36.159 --> 00:51:37.039
Like, try try those things.
00:51:37.119 --> 00:51:38.480
It's it's cool in the car, yeah.
00:51:38.639 --> 00:51:38.960
Yeah.
00:51:40.079 --> 00:51:40.960
All right, this is amazing.
00:51:41.119 --> 00:51:46.800
Last question we ask all of our guests what conventional wisdom or advice do you think is generally bad advice?
00:51:48.480 --> 00:51:52.239
So, I mean, the I mean all kinds of advice can be good or bad advice.
00:51:52.320 --> 00:51:53.440
It all depends on the situation.
00:51:53.840 --> 00:52:14.000
Do you always see like um the main thing is like in terms of um a lot of the good advice that used to be good, like few years back, is now bad advice, is I would say like a lot of the things have changed under everyone's feet on how to build businesses, on how to think about funding, on what makes someone successful, not successful.
00:52:14.079 --> 00:52:15.199
Everything has changed.
00:52:15.440 --> 00:52:17.840
So I would say a lot of the advice that even comes from YC.
00:52:17.920 --> 00:52:24.880
Like one of the examples is like a few years back, you will start in AI any kind of company, you will go after some of the marquee logos.
00:52:24.960 --> 00:52:27.519
You will go say, Okay, I want to get Uber as my customer.
00:52:27.679 --> 00:52:31.280
Because if they are buying your solution, because everyone looks up to them.
00:52:31.760 --> 00:52:33.519
It's a social proof or something.
00:52:33.840 --> 00:52:36.079
Yeah, it's almost like the crossing the chasm curve.
00:52:36.239 --> 00:52:36.960
You have seen that.
00:52:37.199 --> 00:52:39.599
So you have the early adopters and and so on.
00:52:40.079 --> 00:52:41.840
Now, YC still does that.
00:52:41.920 --> 00:52:46.719
They'll go and tell their entire bash to sell to these marquee logos, they're gonna fool around, spend time.
00:52:47.039 --> 00:52:50.400
Interestingly, that curve has now inverted.
00:52:50.559 --> 00:52:52.320
It's invert crossing the chasm.
00:52:52.559 --> 00:52:57.920
In fact, the companies that are benefiting the most from AI are actually more mature companies.
00:52:58.000 --> 00:53:02.159
In fact, they are one of the first ones to adopt because they benefit a lot more.
00:53:02.239 --> 00:53:03.920
They have a lot of average talent.
00:53:04.079 --> 00:53:07.440
Yeah, the models are making them move faster.
00:53:08.800 --> 00:53:16.639
Even like a lot of people at companies which are like Bay Area um startups, um, they already had good engineers.
00:53:16.719 --> 00:53:23.760
So the initially they didn't see enough value because they were anyway smarter, agent, smarter engineers than the A models were, right?
00:53:23.920 --> 00:53:24.320
Yeah, yeah.
00:53:24.559 --> 00:53:24.880
That's right.
00:53:25.039 --> 00:53:26.159
So now maybe that has inverted.
00:53:26.239 --> 00:53:27.840
Now they are adopting agents as well.
00:53:27.920 --> 00:53:30.239
But at least 12 months back that was not the case.
00:53:30.559 --> 00:53:33.440
Right, like the average engineer at, I don't know, Nike.
00:53:33.679 --> 00:53:37.679
Maybe Nike was never able to hire like the most talented engineers.
00:53:37.840 --> 00:53:40.559
And so they really needed these tools really quickly.
00:53:40.880 --> 00:53:41.119
Right.
00:53:41.280 --> 00:53:45.920
Because what AA is doing is like taking the best practices it has learned to an average person.
00:53:46.079 --> 00:53:49.440
So it may not be like, I mean, for example, I can do marketing.
00:53:49.519 --> 00:53:57.840
I may not be the best marketer, but it does level the playing field for me more than someone who's already like a top marketer, they may not see enough value with these agents, right?
00:53:58.079 --> 00:54:08.400
So so the the curve has inverted, and that's a bad advice everyone's still like living by that when you go to the VCs and you want to like look at the quality of the logos, but none of that holds true in building a real business.
00:54:08.559 --> 00:54:10.079
Code Rabbit is one example.
00:54:10.239 --> 00:54:18.239
Now we have great logos, but to be honest, in a year back when you guys came in, you will remember like half our customers were unknown, I don't know, Japanese companies.
00:54:18.320 --> 00:54:19.760
Yeah, but the love was real, right?
00:54:19.840 --> 00:54:21.199
I mean, yeah, absolutely.
00:54:21.440 --> 00:54:23.760
So this is the first time I've seen it's an invert curve.
00:54:23.840 --> 00:54:28.559
Like companies you will usually reach out five years back at CD C or D stage.
00:54:28.719 --> 00:54:30.880
We were closing them in series A or CED.
00:54:31.119 --> 00:54:32.480
Yeah, this is super interesting.
00:54:32.639 --> 00:54:36.320
Thank you, Hajj Dil uh CEO of Code Rabbit for joining us on Summation.
00:54:36.480 --> 00:54:40.159
By the way, I follow you at Harja S Gill on X.
00:54:40.639 --> 00:54:43.840
I definitely encourage our listeners to engage with you there.
00:54:43.920 --> 00:54:44.960
This has been a ton of fun.
00:54:45.119 --> 00:54:46.480
Really appreciate you being on.
00:54:46.800 --> 00:54:47.119
Thank you.
00:54:47.199 --> 00:54:48.000
Thanks for having me on.
00:54:48.400 --> 00:54:49.440
One more thing before we go.
00:54:49.599 --> 00:54:51.440
I read a blog called Summation.
00:54:51.599 --> 00:54:53.280
It's the same as this podcast.
00:54:53.440 --> 00:55:01.679
The blog is about non-obvious idea sharing on business, talent, data, longevity, and random contrarian takes.
00:55:01.920 --> 00:55:06.480
If you like the conversations on this show, you'll probably like the blog.
00:55:06.639 --> 00:55:07.519
It's free.
00:55:07.679 --> 00:55:09.679
New content comes out twice a month.
00:55:09.840 --> 00:55:13.519
You could subscribe at oron.substack.com.
00:55:13.679 --> 00:55:17.039
That's oron.substack.com.