Software Engineer / Agentic Workflows
currentProduction agentic AI for an AI-powered SaaS platform — LLM orchestration, RAG and multi-agent pipelines across a Python/FastAPI backend and a React/TypeScript frontend.
- ›Ship multi-step agentic workflows that turn LLMs into reliable, production-grade automation.
- ›Engineer orchestration, RAG and agent tooling with LangChain / LangGraph plus vector search.
- ›Build guardrails for non-deterministic output: schema-validated results, deterministic re-checks, human-in-the-loop sign-off before anything counts as final.
- ›Add observability, tracing and self-healing so long-running AI jobs stay reliable in production.
- ›Work full-stack: async backends, distributed task queues, real-time streaming UIs.