May 3
Loading video player...
In this episode we build a complete Smart Student Assistant using Multi-Server MCP architecture with Streamable HTTP transport. One AI agent coordinates three independent MCP servers — Calendar, Notes, and Quiz — while OpenAI handles intent detection and study plan generation. We also walk through two real bugs in the code and fix them step by step. ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 🧠 What You Will Learn ✅ How one agent connects to multiple MCP servers simultaneously ✅ How Streamable HTTP transport works in FastMCP ✅ How OpenAI detects query intent and routes tool calls ✅ How AsyncExitStack manages multiple MCP sessions cleanly ✅ How to build a colour-coded Tkinter UI for your MCP agent ✅ Two real Python bugs diagnosed and fixed live ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 📁 Project Structure - calendar_server.py → port 8101 → get_upcoming_events - notes_server.py → port 8102 → list_notes, read_note - quiz_server.py → port 8103 → generate_quiz, get_revision_tip - agent_client.py → coordinator + Tkinter desktop UI ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 🔗 Resources - Source code on GitHub → https://github.com/praveentechlab/mcp-from-zero-to-prod - MCP Specification → https://modelcontextprotocol.io - FastMCP docs → https://github.com/jlowin/fastmcp - OpenAI Python SDK → pip install openai ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 📺 MCP Series Playlist → IPL Insight Agent — Full LLM + MCP Demo in Python | Build an AI Agent with MCP | MCP Series Ep 4 01 — MCP Intro & Hello World: https://youtu.be/v6ZU1psbLQk 02 — Why Async Python Matters | MCP Tools, Resources & Prompts: https://youtu.be/-STdYXBmoIY 03 — Real API Integration | MCP Server as a Bridge: https://youtu.be/IHA-ZalATeU 04 — Full LLM + Agent Demo ← https://youtu.be/VlS1Ccj2_KE ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 🧑💻 WHO IS THIS FOR? - Python developers learning MCP and Agentic AI - Cloud and solutions architects exploring GenAI tooling - Students and educators building AI-powered assistants - Anyone who wants to understand how real multi-agent systems work 👍 If this helped you, like and subscribe for more MCP, Agentic AI, and Cloud content. 💬 Drop your questions in the comments — I reply to every one. ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ #MCP #ModelContextProtocol #AgenticAI #OpenAI #Python #FastMCP #StreamableHTTP #PraveenTechlab