Engineering
AI Engineer — LLM Systems
Mumbai · On-siteFull-time
You will design and ship the AI core of systems our clients run inside their own infrastructure — from model evaluation to the pipelines that make outputs reliable.
What you'll do
- Benchmark open and proprietary models on client data for quality, latency and cost
- Serve models on vLLM and build RAG over pgvector
- Enforce structured, validated outputs (Zod / JSON schema) for downstream systems
- Build evaluation harnesses that catch regressions before clients do
You might be a fit if
- You have shipped an LLM-powered system real users depend on
- You are comfortable with Python and TypeScript
- You care about evals more than demos