Goldman Sachs accelerates with a new GenAI tool, enhancing code generation and efficiency across the firm.
Goldman Sachs is rolling out its first GenAI tool for code generation to thousands of developers, centralizing all proprietary AI technologies on an internal platform. This move, led by CIO Marco Argenti, initially slowed progress but has since increased efficiency.
The centralized GS AI Platform integrates models from OpenAI, Google, and Meta, enabling developers to switch between models for various use cases. This approach allows Goldman to fine-tune models with internal data while complying with regulations. Developers leverage the platform to build applications, such as an AI assistant for investment bankers that searches documents and extracts analysis.
One significant benefit is the platform’s ability to speed up application development. Argenti notes that development time has shortened from months to weeks. The company maintains a focus on safety and responsibility, ensuring all applications have built-in safety measures from the start.
Goldman’s GenAI tool, Microsoft’s GitHub Copilot, boosts developers’ efficiency by 20%. Future projects include a document translation tool and a research summarization application. While GenAI remains a small part of the tech budget, Argenti sees increasing investment in the future without diverting resources from other critical areas.