This solution is designed for testing and experimenting with different .NET frameworks and libraries to build Generative AI applications and AI Agents. It provides a structured environment to explore various AI providers, chat interfaces, and LLM integrations.

• Framework: .NET 9.0
• Dependencies:
  • LangChain (v0.17.0) - Comprehensive LLM framework
  • LangChain.Providers.Ollama (v0.17.0) - Ollama integration
• Description: A conversational AI chat application using LangChain with memory management and Ollama integration
• Features:
  • Multi-turn conversations with memory
  • Template-based prompting
  • Local Ollama model integration (gpt-oss:20b)

• Framework: .NET 9.0
• Dependencies:
  • Microsoft.Extensions.AI (v9.8.0) - Microsoft's AI abstraction layer
  • OllamaSharp (v5.3.6) - .NET client for Ollama
• Description: A streamlined chat application using Microsoft's AI extensions with real-time streaming responses
• Features:
  • Real-time streaming responses
  • Chat history management
  • Direct Ollama API integration

• .NET 9.0 SDK
• Ollama installed and running locally on port 11434
• VS Code with C# Dev Kit extension

1. Build the entire solution:

   dotnet build LLM-demos.sln

2. Run individual projects:

   # LangChain demo
   dotnet run --project LangChainChat-01/LangChainChat-01.csproj
   
   # Microsoft AI Extensions demo
   dotnet run --project Ollama-MS-AI-Extension-Chat-01/Ollama-MS-AI-Extension-Chat-01.csproj

3. Using VS Code Debug: Use F5 or the Run and Debug panel to launch any configured project.

# Navigate to the solution root
cd /path/to/LLM-demos

# Create a new console project
dotnet new console -n YourNewProject-01 -f net9.0

# Add the project to the solution
dotnet sln LLM-demos.sln add YourNewProject-01/YourNewProject-01.csproj

# Navigate to your new project directory
cd YourNewProject-01

# Example: Add common AI packages
dotnet add package Microsoft.Extensions.AI
dotnet add package OpenAI
# or
dotnet add package LangChain
dotnet add package LangChain.Providers.OpenAI

Edit .vscode/tasks.json and add a new task:

{
    "label": "build-yournewproject",
    "command": "dotnet",
    "type": "process",
    "args": [
        "build",
        "${workspaceFolder}/YourNewProject-01/YourNewProject-01.csproj",
        "/property:GenerateFullPaths=true",
        "/consoleloggerparameters:NoSummary"
    ],
    "problemMatcher": "$msCompile"
}

Edit .vscode/launch.json and add a new configuration:

{
    "name": "Launch YourNewProject-01",
    "type": "coreclr",
    "request": "launch",
    "preLaunchTask": "build-yournewproject",
    "program": "${workspaceFolder}/YourNewProject-01/bin/Debug/net9.0/YourNewProject-01.dll",
    "args": [],
    "cwd": "${workspaceFolder}/YourNewProject-01",
    "console": "internalConsole",
    "stopAtEntry": false
}

# Build the solution to ensure everything is configured correctly
dotnet build LLM-demos.sln

# Test the new project
dotnet run --project YourNewProject-01/YourNewProject-01.csproj

• build-langchain: Builds the LangChain demo project
• build-ollama: Builds the Ollama MS AI Extension demo
• build-solution: Builds the entire solution

• Launch LangChainChat-01: Debug configuration for LangChain demo
• Launch Ollama-MS-AI-Extension-Chat-01: Debug configuration for Ollama demo

1. Task naming convention: Use build-{projectname} format
2. Launch configuration naming: Use Launch {ProjectName} format
3. Console setting: Use "console": "internalConsole" for interactive applications
4. PreLaunchTask: Always reference the corresponding build task

When creating new projects, follow this structure:

YourNewProject-01/
├── YourNewProject-01.csproj
├── Program.cs
├── README.md (optional - project-specific documentation)
├── bin/ (generated)
└── obj/ (generated)

• Microsoft.Extensions.AI - Microsoft's unified AI abstraction
• LangChain - Comprehensive LLM framework with chains and agents
• Semantic Kernel - Microsoft's AI orchestration framework

• OpenAI - GPT models (requires API key)
• Azure.AI.OpenAI - Azure OpenAI services
• Anthropic - Claude models
• OllamaSharp - Local Ollama integration
• Google.AI.Generative - Google Gemini models

• Microsoft.SemanticKernel.Plugins.Memory - Vector databases and embeddings
• LangChain.Databases.Chroma - Vector database integration
• Microsoft.Extensions.AI.Abstractions - AI service abstractions

1. RAG (Retrieval Augmented Generation) applications
2. Multi-agent systems with different AI personalities
3. Function calling and tool integration
4. Vector database integration for semantic search
5. Different prompt engineering techniques
6. AI model comparison and benchmarking
7. Streaming vs. non-streaming response handling

• Microsoft.Extensions.AI Documentation
• LangChain .NET Documentation
• Semantic Kernel Documentation
• Ollama Documentation

Feel free to add new demo projects following the established patterns. Each project should:

1. Use .NET 9.0 as the target framework
2. Include proper documentation in comments
3. Follow the naming convention: {Purpose}-{SequenceNumber}
4. Add corresponding VS Code tasks and launch configurations