I need an experienced developer who feels at home in both TypeScript and Python and has already played with the popular open-source “AI agent” stacks. My immediate goal is to stand up a proof-of-concept chatbot that can reason over structured data stored in a SQL database, surface answers in natural language, and be easy for me to extend later. Here is what I’m after: • Pick and wire up an open-source autonomous-agent framework (LangChain, AgentGPT, SuperAGI or a comparable option you trust). • Expose a clean TypeScript interface so the rest of my app can call the agent with a prompt and receive a streamed response. • In Python, configure the agent’s toolset to pull data from my existing Postgres tables, run simple analytic queries, and return concise answers. • Ship working code, a short README, and a one-click launch script (Docker or npm) so I can reproduce the demo on my own machine. Acceptance will be based on: 1. Chatbot reliably answering basic “what/why/how many” questions by hitting the database, not hallucination. 2. Code quality—clear project structure and docstrings/comments where decisions matter. 3. A quick hand-off call or Loom video walking me through setup and how to add new tools or skills. I’m on a tight timeline, so if you’ve done something similar before and can jump in right away, I’d love to see it.