← Cases
United StatesFinance and insurance2025 – ongoing

Goldman Sachs

Developer copilot for software engineering

Goldman Sachs reported a 20% first-year productivity improvement from its developer copilot

WorkingCompany's own claim

Problem

Goldman Sachs sought to improve software-engineering productivity within a tightly controlled banking technology environment.

Approach

Goldman Sachs introduced a developer copilot for software engineers. It reported a 20% productivity improvement in the first year. Its newer Devin coding-agent pilot was a separate initiative.

Results

20%first-year developer-copilot productivity improvement reported in Observer's interview with Goldman Sachs CIO Marco ArgentiCompany's own claim[1]

Takeaway

The reported productivity improvement concerned the developer copilot. It did not establish a result for the separate coding-agent pilot.

Categories

Business function
IT and R&D
Innovation type
Process innovation
AI technology
Generative AI
Evidence
Company's own claim

Sources

Source summaries

[1]ObserverSep 2025

Observer reported a 20% first-year productivity improvement from Goldman Sachs' developer copilot in its interview with CIO Marco Argenti. The interview also discussed the firm's separate Devin pilot and controls for AI-generated software.

Full details are in the original sources.

Added 1 Nov 2026 · Updated 1 Nov 2026
Open case page ↗