Build a Pizza Ordering Agent
summary
Live workshop shows how to deploy a GPT40 model on Microsoft Foundry Agent Service, create a Python agent and thread, and add RAG by uploading documents into a vector store so the agent answers location and menu questions. It exposes a pizza-calculator via function calling and integrates with a live MCP server for menu queries and placing orders. The demo handles real orders and even computes that 150 people need about 20 extra‑large pizzas, and the end‑to‑end flow works.
AI-written summary · about
chapters (12)
- 0:00Welcome & Overview: Building a Pizza Ordering Agent
- 1:54Introduction to the Workshop and Demo Preview
- 6:08What Are AI Agents? Components and Key Concepts
- 9:00Workshop Setup: Requirements and Environment Configuration
- 15:04Deploying Microsoft Foundry Resource and GPT Model
- 18:02Creating Your First Agent in Python
- 28:57Adding System Prompts and Agent Instructions
- 36:06Enhancing Agent Knowledge with File Search and Vector Stores
- 45:57Adding Function Calling: Pizza Calculator Tool
- 52:15Integrating MCP Server for Pizza Ordering
- 57:07Running the Agent and Placing an Order
- 1:02:00Wrap-Up: Best Practices and Open Source MCP Server
related talks
01h 01mAI Agents In-Depth – Function Calling, MCP and Tool Use Under the Hood
32mBuild an AI-Driven Chatbot
45mBuild an agent with Mapbox MCP
40mAI Agents in Practice
36mBuilding AI Agents in .NET with Azure AI Foundry
50mBuilding an entire organization in one line
52mPatterns for Productive Agent-Assisted Programming
01h 58mBuild, Evaluate and Deploy a RAG-Based Retail Copilot with Azure AI