Currently, about a quarter of our pull requests are committed by Devins, and we expect this to exceed 50% by the end of the year.

Each engineer on Cognition's 15-person engineering team works with about five Devins each, enhancing productivity and collaboration.

Our engineering team of 15 uses about five Devins each to help them code and move faster, with Devins merging several hundred pull requests into production every month.

Cognition encourages feedback from users to continuously improve the Devin product and its capabilities.

Cognition went through eight pivots before landing on their current approach with Devin, highlighting the iterative nature of product development.

90% of the code for some products is now written by AI, leading to unexpected bottlenecks in other areas like the merge queue, which had to be completely rearchitected due to the increase in pull requests.

Developers are experiencing a different reality with AI tools like GitHub Copilot, which significantly boosts their productivity.

The development of autonomous coding agents like Devin is a significant step forward in the evolution of software engineering.

The current batch has a notable increase in developer tools, with about 30% more than four years ago, indicating a growing interest in building tools for AI development.