Senior Software Engineer
Software Engineer - Technical Advisor🗓️: 6-month contract (W2)📍: United States (Remote)💰: $150 per hourAn established name in the AI research and safety space, working on frontier AI systems, is on the lookout for Senior Software Engineers. The team is currently focused on understanding how coding agents perform on real, production-grade software engineering tasks, and this hire supports that evaluation work directly.Role OverviewThis role acts as a high-bar technical authority rather than a feature-building engineer. You'll spend your time reviewing model-generated pull requests, working through full agent session logs, and building hard, container-based test problems, using deep code review judgment and clear written rationale to separate genuinely sound engineering from code that only looks right on the surface.ResponsibilitiesAudit model-generated pull requests against real production repositories, documenting every issue found along with its severity and the reasoning behind itReview full coding-agent sessions to work out what the model investigated, verified, assumed, or skippedDesign and build container-based (Docker) technical benchmarks used to test AI modelsWrite clear, original technical rationale explaining why code fails — all written work must be self-authored, no AI text generatorsCollaborate asynchronously with researchers to share findings and refine evaluation criteriaRequired Skills and Experience 8+ years of production engineering experience preferred,Background as a Senior, Staff, or Principal Software Engineer, Tech Lead, or Open-Source MaintainerExperience in production codebases with a strict code review culture — startup, big tech, or open source all countStrong Python and TypeScript experience, with comfort dropping into unfamiliar languages on short noticeSolid hands-on comfort with Docker, git, and the command line for reproducing and debugging issues locallyComfortable working across backend, frontend, APIs, data, testing, or developer toolingStrong written communication — able to explain why code fails, not just how to patch it