
- Artificial Intelligence
Goldman Sachs Deploys AI Engineer Devin, Speeds Dev by 40%
5 minute read

Goldman Sachs AI software engineering pilot delivers 40% faster development times across banking operations
Key Takeaways
- Goldman Sachs pilots autonomous AI software engineer named Devin from startup Cognition, marking the first major bank deployment of AI agents capable of full-stack development tasks
- Hundreds to thousands of AI developers planned as Goldman prepares to scale the program across its 12,000-person development workforce, with early results showing 40% faster delivery times
- Banking industry faces 200,000 job cuts over the next three to five years as AI implementation accelerates across Wall Street firms
Introduction
Goldman Sachs has deployed its newest hire—one that never sleeps, demands no salary, and works around the clock. The investment bank is testing Devin, an autonomous AI software engineer from startup Cognition, in what represents the first major Wall Street deployment of agentic artificial intelligence.
The move signals a fundamental shift from basic AI chatbots to sophisticated agents capable of completing full software development projects independently. Chief Information Officer Marco Argenti confirms the bank plans to onboard hundreds of these AI developers, potentially scaling to thousands depending on performance outcomes.
Key Developments
Devin emerged from Cognition’s labs in late 2023 with claims of being the world’s first autonomous AI software engineer. Demo videos showcased the program completing multi-step coding assignments with minimal human intervention, handling everything from app development to legacy system migrations.
Goldman’s pilot program focuses on tasks that human engineers typically find tedious, particularly updating internal code to modern programming languages. The AI operates under human supervision but executes complex workflows that traditionally require entire engineering teams.
The bank positions Devin as a “developer co-pilot” rather than a replacement, emphasizing augmentation of existing capabilities. Early testing shows the AI can boost productivity by three to four times compared to previous AI tools, with outputs subject to strict auditing and compliance protocols.
Market Impact
Cognition’s valuation doubled to nearly $4 billion following the Goldman partnership announcement, attracting investments from prominent venture capitalists including Peter Thiel and Joe Lonsdale. The startup’s rapid growth reflects broader market enthusiasm for agentic AI solutions.
Tech giants report significant AI code generation adoption, with Microsoft and Alphabet seeing 30% AI-generated code on some projects. Salesforce CEO Marc Benioff notes AI handles up to 50% of work at his company, establishing benchmarks Goldman seeks to match in financial services.
The deployment triggers fresh concerns about job displacement across Wall Street. Bloomberg research projects banks worldwide will eliminate up to 200,000 positions over the next three to five years as AI implementation accelerates.
Strategic Insights
Goldman’s approach represents a calculated shift toward hybrid workforce models where humans and AI collaborate on complex tasks. The strategy focuses on maintaining competitive advantages through proprietary AI integration with trading and risk management systems.
The bank creates internal “AI champions” tasked with identifying effective use cases across business units. This organizational structure drives adoption while ensuring alignment with regulatory requirements and security standards essential in financial services.
Success metrics include 40% improvement in delivery times for standard coding tasks and 15% reduction in post-release bug reports. These early results provide templates for expansion into other operational areas beyond software development.
Expert Opinions and Data
“We’re going to start augmenting our workforce with Devin, which is going to be like our new employee who’s going to start doing stuff on the behalf of our developers,” Argenti told CNBC. He envisions scaling from hundreds to potentially thousands of AI agents depending on use case development.
Belinda Neal, Goldman’s chief operating officer of core engineering, emphasizes the scale of transformation. “For us, this year is a story of scale — scale of adoption, scale of use cases,” she explains, highlighting the firm’s commitment to comprehensive AI integration.
The technology addresses practical engineering challenges while maintaining human oversight. “Engineers are going to be expected to have the ability to really describe problems in a coherent way and turn it into prompts,” Argenti notes, emphasizing supervision over competition with AI systems.
Conclusion
Goldman Sachs establishes itself as the first major bank to deploy autonomous AI software engineers, setting precedents for Wall Street’s technological evolution. The pilot program demonstrates measurable productivity gains while maintaining the regulatory compliance and security standards essential to financial services.
The initiative represents a broader transformation from basic AI assistance to sophisticated agent-based automation. As results continue to validate the hybrid workforce model, other financial institutions face pressure to accelerate their own AI adoption strategies to maintain competitive positioning in an increasingly automated industry.