About This Video

In this comprehensive tutorial, viewers will learn how to build an AI-powered dinner decision agent using a combination of Microsoft Semantic Kernel, MongoDB Atlas (with Vector Search), C#, and OpenAI. Led by MongoDB Developer Advocate Luce Carter, the video demonstrates how to set up a .NET application that helps users decide what to cook—or suggests nearby restaurants—based on the ingredients they have on hand and their preferred cuisine.

The walkthrough covers essential topics including configuring Semantic Kernel plugins, managing ingredient lists, building custom prompts, and integrating OpenAI for advanced reasoning. You'll explore how to leverage MongoDB Atlas's Vector Search to match restaurant recommendations to user preferences and see practical techniques to bring generative AI into your .NET applications.

Key Points

• How to set up and use Microsoft Semantic Kernel with C# for AI agent applications.
• Integrating MongoDB Atlas, including vector search for semantic restaurant recommendations.
• Building and leveraging custom plugins and prompts to make dynamic, context-aware decisions.

Prerequisites

• Access to a free MongoDB Atlas M0 cluster with sample data loaded.
• An OpenAI account and API key (free tier sufficient).
• .NET 9 SDK installed on your machine.
• Basic familiarity with C# and .NET console applications.

Resources Mentioned

• Written Tutorial: How to Build an AI Agent with Semantic Kernel, MongoDB Atlas, C# and OpenAI
• Code Snippets: GitHub Gist with all snippets
• Microsoft Learn Course: Getting Started with Semantic Kernel
• Vector Search Documentation: MongoDB Atlas Vector Search
• MongoDB Developer Center: Explore more on MongoDB Developer Center
• Subscribe: MongoDB YouTube Channel