Job Description
About the Role
Build and deploy production-grade AI agent systems on our enterprise GenAI platform. You'll own the full lifecycle — from agent design to cloud deployment — with a pragmatic bias toward simplicity and measurable ROI.
What You'll Do
- Build multi-agent systems using LangGraph, LangChain, Google ADK, and PyVegas following A2A and MCP protocols
- Expose AI logic as FastAPI services with async task handling and streaming
- Own deployments on Cloud Run — Dockerfiles, Cloud Build CI/CD, Cloud SQL, Secret Manager, canary rollouts
- Instrument agents with LangSmith — tracing, evaluation pipelines, prompt versioning, and CI-gated regression tests
- Apply context engineering discipline: RAG re-ranking, token budget management, structured tool responses
- Default to deterministic code; escalate to agents only when the problem genuinely requires reasoning
Required
What You Bring
- 4+ years Python engineering with production deployments
- 2+ years with LangChain / LangGraph (v1.x) — stateful agents, graph compilation, checkpointing
- LangSmith — tracing, evaluations, prompt hub, CI integration
- FastAPI — async endpoints, middleware, background workers
- Docker + GCP — Cloud Run, Cloud Build, Artifact Registry, Cloud SQL, Secret Manager
- PostgreSQL for agent state, task queues, and LangGraph persistence
Preferred
- PyVegas or equivalent enterprise LLM platform wrapper
- Google ADK, DSPy, LiteLLM, Vertex AI / Gemini
- A2A / MCP protocol experience
- New Relic or Galileo for GenAI observability
Tech Stack
LangGraph LangChain LangSmith PyVegas Google ADK DSPy LiteLLM FastAPI Cloud Run Vertex AI PostgreSQL Docker MCP A2A
Qualifications
Masters
Range Of Year Experience-Min Year
8
Range Of Year Experience-Max Year
10
