Goldman Sachs Integrates AI Engineer 'Devin'

Goldman Sachs is integrating the autonomous AI software engineer, Devin, into its development workflow. The move signals a significant step in enterprise adoption of agentic AI, where complex coding tasks are operationalized with substantial autonomy.

- Devin was created by Cognition AI, a startup valued at nearly $4 billion, which is backed by prominent investors like Peter Thiel. - On the SWE-bench benchmark, which uses real-world GitHub issues to test performance, Devin was able to resolve 13.86% of issues end-to-end without assistance, a significant increase from the previous state-of-the-art score of 1.96%. - Unlike AI-powered code completion tools such as GitHub Copilot that assist developers, Devin is designed to operate as an autonomous agent, capable of handling entire development projects from coding and debugging to deployment with minimal human intervention. - Goldman Sachs is piloting Devin with its 12,000-person engineering team, initially deploying hundreds of instances with the potential to scale to thousands, to handle tasks like legacy code management and debugging. - The AI agent operates within a sandboxed environment that includes a shell, code editor, and browser, allowing it to mimic a human developer's workflow. - Despite its impressive benchmarks, some developers are skeptical, pointing out that Devin's demonstrated successes are often on less complex tasks and that it can struggle with the context of large, existing codebases. - Goldman Sachs' CIO, Marco Argenti, has referred to Devin as a "new employee" that will augment the existing workforce, envisioning a future "hybrid workforce" model where human engineers oversee and collaborate with AI counterparts.

Get your own daily briefing

Scout delivers personalized news, insights, and conversations tailored to your role and industry.

Download on the App Store

Shared from Scout - Be the smartest in the room.