What Data from 20 Million Pull Requests Reveal About AI Transformation
AI Engineer Code 2025 · 17:57
Software Engineering Intelligence
Jellyfish builds a software engineering intelligence platform that helps R&D leaders connect engineering work with business priorities. It combines data from engineering tools with business context to support planning, delivery and performance decisions. Its Allocation framework measures how engineering effort is distributed across features, infrastructure, bug fixes and other priorities in full-time-equivalent units, helping leaders assess capacity and investment. The AI Impact Dashboard, introduced in 2024, extends that measurement approach to the impact of AI coding tools.
Founded in 2017 by Andrew Lau, David Gourley and Phil Braden, Jellyfish grew out of its founders’ experience managing engineering organizations and communicating their business impact. Lau remains CEO. Alongside the platform, Jellyfish Research studies how AI changes software engineering. Its H1 2026 review examines token spending, agent autonomy and codebase architecture, distinguishing increased AI consumption from improved output. Observational analysis links sustained investment in agent context files with higher pull-request throughput, making the environment around a coding agent a focus of its research.
In its May 2026 announcement, Jellyfish reported serving more than 700 companies, including DraftKings, Keller Williams and Blue Yonder. Its financing history includes a 2020 Series A followed by Series B and C rounds in 2021 and 2022, bringing total funding to $114.5 million.
AI Engineer Code 2025 · 17:57
Affiliations reflect their AIE appearances, not necessarily current employment.
Arcolano identifies repository architecture and context limitations as constraints on the productivity gains teams can obtain from AI coding tools.
Affiliations reflect each recorded session, not necessarily current employment.