Huge shoutout to the team on the insanely fast execution on this. Lots more to ship.

18 million miles traveled, 400 red hoodies, one 260-digit number, and 130+ former founders.

We asked some of the people building Devin to tell us how they got here and what life at Cognition is really like. Here’s what they shared.

BREAKING: @Mclader has joined Cognition as Head of Cognition Industries

Update: I am joining the @cognition team as the new Head of Industries, after selling Native Markets’ USDH to Coinbase.

I've known @ScottWu46 since he was a beta tester of Uniswap products and have been bullish on Devin since the first demo. I couldn’t be more excited to drive iring entrepreneurs who've been building on hard mode to join a team that will transform how every vertical builds software.

If you’re a founder or operator who has shipped and sold products in the real world, and you have a vision for what’s now possible in your industry with AI, shoot me a DM. I’d love to chat.

Professional line dancers, Melee champions, a pyrotechnician who programs fireworks shows with Devin …

Join the fun: cognition.com/careers

New episode of The Marketing Engineer Podcast with @bcamhi, SVP of Marketing at @cognition.

Cognition just passed $1 billion in revenue, but when Brandon joined its marketing team was still early.

So the team hired a Marketing Engineer. Within a week and a half, all of re managed by AI, and list uploads from the field team were too.

In this episode, Brandon breaks down how he's building the marketing team at Cognition with an AI first approach.

I got sick of reviewing code. So I made /walden-review.

Its a skill auto-generated against my last 6 months of Github comments and agent sessions. It learned what I look for and even my comment style: “blunt, occasionally profane”.

I always run it before ever reading a PR design/codepattern/architectural problems most bug-catcher miss. FYI: This method burns a lot of tokens

First, I asked Devin to collect my last 6 months of GitHub PRs and Devin sessions, collect every comment I left or message I sent, filter out unimportant stuff (“lgtm”), and build a hillclimbing set of PRs with a training and eval split.

Then I had it recursively improve a skill to catch as many of the issues I’d flagged on this hillclimbing set as possible, while leaving at most 2x as many comments as I did. In the process, I found that it performed much better once I told it to add “subagent sweeps” to the skill: each sweep spawns a Fable/Astra subagent with specific responsibilities to look for certain things I would flag.

There’s some nuance in how to grade the results. I personally found that some decisions require product context and aren’t reasonable to expect the AI to catch. So I split out the “code concerns” from “product concerns” and mostly tried to measure it against the code concerns, with the expectation that I would make the product calls. For product calls, I use a separate skill that generates a Markdown explainer of the important decisions in a PR. I also use an AI grader to make the judgement call about which bugs it flags are “close enough” to the bug reports I would’ve made.

I ran the hillclimbing overnight. Because my approach ran multiple levels of graders, idea generators, and review swarms, this takes a lot of compute. I came back to find that the hillclimbing used 91 subagents, 559 grandsubagents, and 1,883 great-grandsubagents.

On the held-out eval set, it catches about 70% of the code issues I’d flagged. Now, whenever I get a massive PR, I just throw this skill at it, and let it auto-fix a bunch of the issues I would’ve caught. The skill also separates issues into “auto-fix” and “requires Walden’s approval,” so I sometimes come back to a list of decisions I need to make.

I’m pleasantly surprised that it catches a lot of the things I get most annoyed by. On its own, for example, it learned to double check assumptions about external systems, it learned to aggressively consolidate with existing code, and it learned to remove dead code.

I’ve also generated versions of this for other people on the team who write a ton of code. Honestly, I think there’s a lot more someone could do with this pattern. I haven’t tried training it on ALL of our human review comments yet, but that might be interesting. I also suspect there’s similar things people can do along these lines to catch security vulnerabilities before they ship.

I’ve attached the final skill in the thread as well as a meta-skill I use to generate the same thing for other people on the team.

It’s crazy how many product decisions at @cognition boil down to: “What would a cracked teammate do?” 🤯

Another example…

A-players are really proactive. Devin watches for issues across Slack, etc. and then jumps in to help resolve them.

What could you get done this week if you had an unlimited team of A-players?

We’re building cracked marketing infrastructure at Cognition and we’re going to run the entire team on Devin

One thing that gets lost in the buzz around @cognition: it’s such a fun place to work.

Insane dedication to craft.
Default optimism & trust.
Electric energy.
Little red tape.

Haven’t seen anything quite like it.

wow… devin is actually great now wtf. the ads were spot on

Welcome @alexstamos to Cognition.

Hypergrowth is the single most fun thing in the world. I've seen it a few times but NEVER at this speed 🚀

Come join us! cognition.com/careers

One striking thing about this milestone is that we got here while continuing to lower our prices for customers.

Devin Fusion, SWE-2, and our ROI guarantee align incentives for the long term. It's an approach I'm really proud we take.

Grateful for all our customers!

Cognition has crossed $1B in annualized revenue run rate.

This milestone belongs to our customers. Here's how a few of them are building with Devin.

A big part of the reason I joined Cognition is @ScottWu46's and @russelljkaplan's buy-in to go huge in marketing...can't wait to build with them and the rest of the team 🚀