What we’re building
Respan is building the self-driving observability and evals platform for AI teams, used by 60+ YC companies and hundreds of AI teams.
The role
You will work on the systems that make Respan intelligent: automated evaluations, prompt optimization, model routing, and AI-powered analysis of traces and logs. This is deeply technical AI engineering, not just API calls. We are looking for someone who understands LLMs from the inside out and can build reliable, production-grade AI systems that thousands of teams depend on daily.
What you’ll do:
- Design and build AI-powered features including eval frameworks and prompt optimization
- Develop LLM-as-judge pipelines for automated quality assessment
- Build and maintain model routing, caching, and fallback systems
- Experiment with new model capabilities and translate them into product features
- Collaborate with customers to understand AI workflow pain points and build solutions
What you must have:
- Strong software engineering skills with Python and TypeScript
- Hands-on experience building with LLMs in production (not just prototypes)
- Deep understanding of prompt engineering, eval methodologies, and model behavior
Strong plus:
- Experience with fine-tuning, RAG, or agent frameworks
- Background in ML infrastructure or MLOps
- Familiarity with observability systems or developer tools
- Contributions to open-source AI/ML projects