Reproduce “Pomona: Continuous Code Quality Improvement via Small, Agentic Pull Requests at Bloomberg” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

David Williams, Angelos Evripiotis, Serkan Kirbas, Harry Morgan, Sergey Magidovich, Peter Wainwright, Federica Sarro

Pomona organizes code maintenance through a prioritized backlog and two agent skills: one finds and ranks fixes, while the other implements the top task, validates it, and opens a pull request for human review. In a three-month deployment at one Bloomberg team, 32 of 39 pull requests were merged (82.1%); this experience does not establish better code quality or productivity.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.