All open roles
EngineeringFull-time·3+ years
Evaluation Engineer
Own the eval contracts that ship with every client deployment. Build the tooling that makes evaluations reproducible, fast, and trustworthy.
Dhaka, Bangladesh · Remote
What you will do
- Design and ship eval suites tied to client-specific safety contracts.
- Build CI integrations so evals run on every model checkpoint.
- Triage regressions and route them to the right owner.
What we need
- Strong Python. Comfortable with distributed eval orchestration.
- Care about reproducibility and observability as much as the result.
- Experience working closely with researchers.
Nice to have
- Prior work on LLM eval frameworks.
- Background in software testing or QA infrastructure.