This directory contains all the code samples for the Generative AI for Beginners .NET workshop. The samples are organized into four main categories based on their purpose and complexity.

samples/
├── CoreSamples/          # Core AI techniques and fundamental concepts
├── MAF/              # Microsoft Agent Framework samples
├── PracticalSamples/     # Real-world practical applications
└── AppsWithGenAI/        # Applications created with GenAI assistance

Location: samples/CoreSamples/
Associated Lessons: Lesson 02 & Lesson 03

This folder contains fundamental samples demonstrating core Generative AI techniques with .NET. These samples cover the essential building blocks for AI-enabled applications and are referenced throughout Lessons 02 and 03.

• Chat & Completions - Basic text generation and conversational AI using various providers (Azure OpenAI, Ollama) and frameworks (Microsoft.Extensions.AI)
• Function Calling - Extending AI capabilities by allowing models to call custom functions and external tools
• Retrieval-Augmented Generation (RAG) - Integrating external knowledge stores with AI models using different vector databases (in-memory, Azure AI Search, Qdrant)
• Vision & Audio - Image analysis, speech-to-text, and real-time audio conversations
• Agents - Basic agent implementations with custom functions and OpenAPI integrations
• Image & Video Generation - Creating visual content with DALL-E and Sora models
• Model Context Protocol (MCP) - Integrating with external services using MCP
• Local Model Runners - Running AI models locally using AI Toolkit, Docker, and Microsoft Foundry Local

This folder includes over 40 samples demonstrating different techniques, providers, and integration patterns.

Location: samples/MAF/
Associated Lesson: Lesson 04

This folder contains samples demonstrating the Microsoft Agent Framework for building sophisticated multi-agent AI systems.

• Basic Samples - Single and dual-agent workflows demonstrating fundamental agent creation and orchestration patterns
• Advanced Multi-Provider Samples - Complex workflows combining multiple AI providers (Azure OpenAI, Ollama) in sophisticated agent orchestrations
• Integration & Web Samples - Web-based chat applications with Blazor UI, middleware patterns, and MCP integrations for tool usage (e.g., image generation)

This folder includes 12 samples ranging from simple single-agent scenarios to complex multi-agent systems with persistent state and web interfaces.

Location: samples/PracticalSamples/
Associated Lesson: Lesson 03

This folder contains practical, real-world samples demonstrating AI integration patterns in production-ready scenarios.

A comprehensive .NET Aspire MCP Sample showcasing the Model Context Protocol (MCP) integration with a complete multi-project solution including:

• .NET Aspire orchestration for managing distributed services
• Blazor-based chat interface for user interactions
• ASP.NET Core backend integrating with MCP servers
• Shared service configurations and defaults

Run the sample:

cd samples/PracticalSamples/src
dotnet run --project McpSample.AppHost/McpSample.AppHost.csproj

Location: samples/AppsWithGenAI/
Associated Lesson: Lesson 05

This folder showcases complete, production-quality applications created with the assistance of Generative AI tools like GitHub Copilot Agent, demonstrating "vibe coding" - building real software through AI-assisted development.

• SpaceAINet - A retro space invaders console game with AI-powered enemy behavior that can analyze game state in real-time and make strategic decisions using Azure OpenAI or Ollama
• HFMCP.GenImage - An image generation application using Hugging Face's MCP Server, demonstrating multi-modal AI capabilities with .NET Aspire orchestration
• ConsoleGpuViewer - A GPU diagnostics and monitoring tool that integrates with local AI models for intelligent hardware analysis

Each application is fully functional and can be run independently with proper AI provider configuration.

• .NET 10 SDK or later
• AI Provider Access (at least one):
  • Microsoft Foundry endpoint (with managed identity or API key)
  • Ollama installed locally

Note: Use the appropriate path separator for your platform:

• Windows (CMD/PowerShell): Use backslashes \ (e.g., samples\CoreSamples\BasicChat-01MEAI)
• Linux/macOS/Git Bash/WSL: Use forward slashes / (e.g., samples/CoreSamples/BasicChat-01MEAI)
• GitHub Codespaces: Always use forward slashes / (runs Linux environment)

cd samples/CoreSamples/BasicChat-01MEAI
dotnet run

cd samples/MAF/MAF01
dotnet run

cd samples/PracticalSamples/src
dotnet run --project McpSample.AppHost

cd samples/AppsWithGenAI/SpaceAINet/SpaceAINet.Console
dotnet run

• Main Course README

• Setup Environment

• Microsoft.Extensions.AI Documentation

• Microsoft.Extensions.AI Documentation

• Microsoft Agent Framework

• Web chat with persisted conversations - MAF-AIWebChatApp-Persisting demonstrates per-user AgentThread serialization, storage, and resume across restarts