Turing lists this role as Global - Remote, so applicants from any country can apply (always check the official posting for any local employment limits).
Before an AI lab tells the world its coding agent got better, someone has to prove it with thousands of trials that can be trusted. Turing is hiring a senior engineer to build the infrastructure that makes those evaluations possible.
About Turing
Turing is a San Francisco company that works with frontier AI labs to produce high-quality datasets, reinforcement learning environments and research benchmarks. It also works with large enterprises in finance, life sciences, healthcare, retail and automotive.
About the Role
- Design and ship features from data model and API to background workers and the screens users work in
- Improve how agents and models are evaluated, including trials in sandboxed environments and scoring at scale
- Improve how benchmark tasks are quality-checked, with verifiers and rubric-based or LLM-as-a-judge grading
- Add new models, agents and benchmarks without writing a special case for each
- Keep the system fast and reliable as usage grows
- Review teammates' pull requests against the written engineering guidelines
About You
- Typically 5 or more years building and running production backend systems
- Strong Python, with Django or a similar framework, PostgreSQL and a task queue such as Celery
- Hands-on experience in LLM evaluation, beyond calling a model API
- Good judgment about task quality, such as spotting weak or gameable verifiers
- Clear writing and comfort with a TypeScript and React front end when needed
Nice to Have
- Experience running agents in sandboxed or containerised environments such as Docker or Kubernetes
- Experience with LLM-as-a-judge grading and with coding agents such as Claude Code, Codex or OpenHands
- Open-source contributions to evaluation or infrastructure projects
What Turing Offers
The posting does not list a salary or benefits. The interview process has four steps: a conversation about a system you built, a practical coding exercise, a discussion of a flawed benchmark task and a team conversation.
If this sounds like your kind of problem, apply on Turing's official careers page: Apply for the Turing Senior Software Engineer role