Typical responsibilities
- Build retrieval-augmented generation (RAG) over your documents, knowledge bases and data.
- Design prompts, tool calls and multi-step agents for real business workflows.
- Create evaluation sets and automated tests to track accuracy and regressions.
- Add guardrails for privacy, security and off-topic or unsafe answers.
- Fine-tune or adapt models when prompting alone is not enough.
- Optimize latency and cost with caching, model routing and smaller models.
- Monitor live conversations and improve the system from real usage.
