Contact person: LangGraph Framework Implementation

Phone:Show

Email:Show

Location: London, United Kingdom

Budget: Recommended by industry experts

Time to start: As soon as possible

Project description:
"I’m building an advanced, multi-agent LLM system and want the framework layer nailed down first. LangGraph will be the backbone, so I need someone who already speaks its event-driven, graph-based language fluently.

The immediate goal is to design and stand up a clean LangGraph scaffold that can later plug into function-calling tools, RAG pipelines, and vector stores such as pgvector or Weaviate. You’ll translate a high-level workflow spec (nodes, edges, state, failure paths) into working code, set up unit tests for each prompt or tool node, and document the pattern clearly so additional agents can be dropped in without rewiring the whole graph.

Please show past work where you:
• built or contributed to a LangGraph, or a comparable graph-style orchestration (AutoGen, CrewAI, etc.)
• demonstrated solid prompt-level testing, retry logic, or memory management for multi-turn agents

I already have keys for OpenAI and local vLLM, plus a Dockerized dev environment; you can extend or tweak it as needed. If you’ve optimized graph execution on Ray or Modal before, mention it—it will help us iterate faster later.

Deliverable: a repo (or PR) containing the LangGraph implementation, test suite, and concise setup docs. Once merged and smoke-tested, we’ll roll straight into tool orchestration and eval harness work." (client-provided description)