The Age of Async Agents — Cognition's Walden Yan & OpenInspect's Cole Murray

**Swyx** (0:03)
All right, we're in the studio with Walden Yan, co-founder of Cognition, CPO.

**Walden Yan** (0:08)
Yeah, happy to be here.

**Swyx** (0:09)
Cool title. Yes, and coiner of context engineering.

**Walden Yan** (0:14)
Yes. Although I think there are many people who use the terms of various ways beforehand, but I did find that people both internally and externally enjoyed the upgrade from front engineering or model wrapping into maybe a more thoughtful way to build agents.

**Swyx** (0:33)
Yeah, for those who haven't caught up on that, I have on screen the Don't Build Multi-Agents post which you should read on and we might refer to, and Cole Murray who created OpenInspect.

**Cole Murray** (0:43)
Created by me.

**Swyx** (0:43)
Okay, so let's talk about it. Everyone is building their own Devins. What's going on?

**Cole Murray** (0:51)
Yeah, so I think the engineering world is kind of waking up to this idea of background agents, Cloud agents, whatever you'd like to call it. I think we saw a shift around the December timeframe of 2025, where the models Opus 4.5 and Gpt 5.2, they reach a capability where we moved away from kind of handholding the model and being able to actually, more or less, autonomously drive the model. What I mean by that is that we could pretty much go from a specification to a completed pull request, assuming the spec was good enough, with very little friction.
That paradigm alone, I think, changed a lot of how we interact with agents and kind of opened this world where background agents became more practical.

**Swyx** (1:41)
I think for Claude, everyone experienced this in December, but I feel like there was just this increasing ramp, right? Like, there was this moment, which was, I think, Sonnet 3.7, where like you guys rewrote Devin in one night.

**Walden Yan** (1:55)
Yes. Yeah.

**Swyx** (1:56)
So describe 2025 or how it felt from your side.

**Walden Yan** (2:00)
In retrospect, we always thought it was ramping up, but then even now, over the last three, four months from today, it's been ramping up even faster. So it's almost funny to be talking about how big of a leap Sonnet 3.7 was. And we honestly, a lot of it was stripping out parts of Devin that were no longer needed with that jump in of intelligence. But I also just think that a lot of the recent leaps, especially you look at models like Opus and latest Gpt models, they are reaching levels of autonomy where people are actually fighting that they actually can't just be hands off. And people who were once debating, oh, do I need to be in the weeds? Is my model in the eye to eat? Can I just completely move it off into the cloud?
That's a more serious conversation. We've seen that in all of our growth charts. Internally, there's this funny graph where our usage has of PRs, our merged PRs, has grown 7x since.

**Swyx** (2:56)
Yes.

**Walden Yan** (2:57)
I forget what it looks like.

**Swyx** (2:58)
I think Dave maybe tweeted that.

**Walden Yan** (3:00)
Yeah.

**Swyx** (3:00)
Yes.

**Walden Yan** (3:01)
It grew like 7x over the last, I think it was like two months, three months, something like that. And then you see our engineering headcount growth. It's gone up by 10% or something.

**Swyx** (3:10)
Like, we were afraid to release this. So, this is Devin commit percentages on all Devin repos was 16% in January and now 80% in March.

**Walden Yan** (3:22)
Yeah.
It's like, it's a big shift right now. And so, it makes sense that a lot of people are now thinking about, you know, buying Devin, but also maybe like, you're trying to build their own. And there's lots of, I have a lot of fun building Devin, so I can see why other people would want to build their own cloud agents as well. Yeah. Well, maybe it's good to hear, like, what initially inspired you to try to build OpenInspect?

**Cole Murray** (3:49)
Yeah. OpenInspect came about through primarily my clients observing how they were using tools like ClaudeWeb, OpenAI's codecs at the time, and seeing some of the friction that they were having with it. Primarily, the ClaudeWeb was being used through Slack. And a big issue they ran into is that the sessions that were launched were specific to whoever called it via Slack. And so if a PM was the one who invoked the session, and they would then go to past context engineering, engineering can't see the session. And that in itself was kind of a deal breaker because the PM said, hey, engineering, can you jump in? But there's nothing to jump in on unless they're copy pasting out or the single response that came back.