demo 0:60
How I Built an AI Agent Over an Existing Backend (Using MCP, Without a Rewrite)
Quick demo showing how to build an AI agent over existing backends using Model Context Protocol, without needing a complete rewrite. Built with FastAPI, FastMCP, and Streamlit.
Overview
A short demonstration of MCP Agent Chat - an AI agent for order management powered by Model Context Protocol. Built with FastAPI + FastMCP + Streamlit.
What’s Shown
- MCP Agent Service running (Backend on port 8080)
- Agent Service running (Port 3000)
- Interactive chat interface
- Order management capabilities:
- Check order status
- Show latest order
- Cancel orders
- Place new orders
Tech Stack
- Backend: FastMCP + FastAPI
- Frontend: Streamlit
- AI: Model Context Protocol (MCP)
- Agent Framework: Custom MCP Agent Service
Key Features
- Natural language order management
- Real-time backend integration
- Safe AI agent operations
- Production-ready architecture
Watch on YouTube
Video Thumbnail:
▶️ Watch Short: Building AI Agent with MCP
Related Resources:
Tags
AIMCPAgentsBackendFastAPIStreamlit