Own the agentic layer of an on-premise AI platform: build multi-step LLM agents on LangGraph (planning, tool use, human-in-the-loop), RAG pipelines, prompt engineering with structured outputs, and evaluation/prompt-injection defences. All models are self-hosted on the company's own hardware — Python, FastAPI, vLLM, Qdrant, OpenSearch, Postgres, Kubernetes.
Sign in to see your match