Staff Software Engineer, Copilot Code Review

About GitHub: GitHub is the world’s leading platform for agentic software development — powered by Copilot to build, scale, and deliver secure software. Over 180 million developers, including more than 90% of the Fortune 100 companies, use GitHub to collaborate, and more than 77,000 organisations have adopted GitHub Copilot.

Locations: In this role you can work from Remote, United States

Overview

AI writes code faster than teams can review it, and review is now the bottleneck. Getting an agent to review as well as, or better than, a great engineer who knows your codebase is one of the most important unsolved problems in AI-assisted development.

Copilot code review is GitHub's answer, and more developers get reviews from it than from any other AI reviewer, increasingly on code that agents wrote. Getting review right is hard to define and harder to measure, and no single number settles whether reviews got better. We think the AI products that win are the ones with the best evals, and we're building the best, offline and online.

We run at GitHub scale with an on-call rotation, and we're a fast-growing team competing to build the best AI code reviewer in the world.

As a Staff Software Engineer here, you'll take the problems nobody has answered yet and ship the answer: when a review should stay quiet, or what the agent should learn about a codebase instead of being told. You won't be handed a spec. Other engineers will follow the calls you make, but you lead by building, not by writing the plan. Responsibilities: * Own product problems that arrive without an answer and ship the answer end to end. * Advance how the agent reviews: the context it uses, what it remembers, and how it improves over time. * Decide what a good review looks like for a class of code, then build the evals that hold the product to it. * Make the quality, latency, and cost tradeoffs that decide how deep a review goes. * Improve review quality in aggregate and quantify the gain,