Full-Stack AI Engineer/Architect: Build Multi-Agent Systems, MCP Servers & Data Clean Rooms

Project Overview: Seeking a hands-on technical co-pilot to build advanced AI systems completely from scratch while actively teaching me engineering mechanics, code architecture, and deployment steps. Phase-by-Phase Execution Plan: Phase 1 (Infrastructure): Initialize local environments, GitHub repositories, CI/CD pipelines, secret management, and core infrastructure templates. Phase 2 (Project Voyager): Code Python (FastAPI) mock APIs for inventory, dynamic pricing, and user loyalty. Build stateless Model Context Protocol (MCP) servers using Python/TypeScript SDKs to expose APIs as tools. Implement LangGraph orchestration for ReAct loops, state graphs, and tool-calling constraints. Deploy MCP servers serverlessly (Cloudflare/AWS) and test end-to-end chat flows. Phase 3 (Data Clean Room): Provision Snowflake and mock zero-party CRM/transaction tables. Write secure data-matching/privacy-compliant query pipelines using Snowpark/Python. Build a real-time attribution API gateway for campaign metrics. Phase 4 (Architecture Walkthrough): Review every line of code to master data flows, latency mitigation, state management, and security boundaries. Ideal Candidate Profile: Production experience with multi-agent frameworks (LangGraph) and MCP. Strong backend engineering skills in Python, FastAPI, SQL, and Snowflake.