--- title: Building Intelligent .NET Applications: From AI Features... description: A workshop-style talk walks through building AI-powered .NET applications, starting with Microsoft.Extensions.AI (MEAI) as an abstraction layer over LLM... canonical: https://daily.dev/posts/building-intelligent-net-applications-from-ai-features-to-agentic-system-rtromjgwu twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Building Intelligent .NET Applications: From AI Features to Agentic System | daily.dev og:description: A workshop-style talk walks through building AI-powered .NET applications, starting with Microsoft.Extensions.AI (MEAI) as an abstraction layer over LLM... og:url: https://daily.dev/posts/building-intelligent-net-applications-from-ai-features-to-agentic-system-rtromjgwu og:image: https://api.daily.dev/og/posts/RTroMjgWU.png og:image:alt: Building Intelligent .NET Applications: From AI Features to Agentic System og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Building Intelligent .NET Applications: From AI Features to Agentic System **[.NET Blog](https://daily.dev/sources/dotnet)** · 77 min read · 0 upvotes · 0 comments ## Summary A workshop-style talk walks through building AI-powered .NET applications, starting with Microsoft.Extensions.AI (MEAI) as an abstraction layer over LLM providers like OpenAI, Azure OpenAI, and Ollama. It covers structured output via JSON schema enforcement, chat state management, and then progresses to retrieval augmented generation (RAG): chunking documents, generating embeddings, and storing vectors in databases like Quadrant, SQL Server, Redis, or Azure AI Search. A live demo shows the AI web chat project template with Aspire orchestration, Microsoft Extensions for Data Ingestion (MEDI) for PDF processing via MarkItDown, and Quadrant's dashboard. The session closes with the GitHub Copilot SDK for building tools that reuse a developer's existing Copilot subscription instead of managing separate LLM backend credentials, plus a teaser for the newly released C# SDK for Model Context Protocol (MCP), enabling Copilot to call into custom company tools. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is Microsoft.Extensions.AI (IChatClient) and how does it help connect to different LLM providers in .NET? Microsoft.Extensions.AI provides the IChatClient abstraction, a single .NET interface for talking to any large language model backend instead of learning each provider's own SDK. Implementations exist for OpenAI, Azure OpenAI, Microsoft Foundry, and Ollama, plus community packages on NuGet for providers like Google Gemini. Swapping providers is typically as simple as changing the connection string and model name. _Developers comparing LLM providers for .NET apps can track ecosystem updates like this on daily.dev._ ### How does structured output work with Microsoft.Extensions.AI in .NET to avoid parsing broken JSON from an LLM? Structured output lets you request a response as a specific C# type (for example a record with typed properties), and Microsoft.Extensions.AI sends a JSON schema to the model behind the scenes so the response is enforced to match that shape and deserialized directly into the type. This avoids manual text parsing and issues like models wrapping JSON in markdown or extra commentary, though how well a model honors the schema still depends on the specific model and provider. _Teams building typed AI responses in .NET can follow guidance like this on daily.dev while shipping features._ ### What is the Microsoft Agent Framework and how does it relate to Semantic Kernel and AutoGen? The Microsoft Agent Framework is the unified successor to Semantic Kernel and AutoGen, now generally available (no longer just a release candidate). It wraps an IChatClient together with tools, context, and instructions into a componentized agent, supports multi-agent orchestration with handoffs to pass context between agents, uses the A2A (agent-to-agent) protocol for communication, and can be paired with an agentic UI layer called AGUI. _Developers evaluating multi-agent frameworks for .NET can keep up with Agent Framework changes via daily.dev._ ## Similar posts on daily.dev - [The Unexpected AI Stack: C\# \+ .NET \(Part 3\)](https://daily.dev/posts/the-unexpected-ai-stack-c-net-part-3--silxqutsc) · ITNEXT · 0 upvotes · 0 comments - [The Core Building Blocks Explained](https://daily.dev/posts/the-core-building-blocks-explained-1o2g8pooq) · .NET Blog · 2 upvotes · 0 comments --- Tags: [#.net](https://daily.dev/tags/.net), [#mcp](https://daily.dev/tags/mcp), [#rag](https://daily.dev/tags/rag), [#vector-search](https://daily.dev/tags/vector-search) [View this post on daily.dev](https://daily.dev/posts/building-intelligent-net-applications-from-ai-features-to-agentic-system-rtromjgwu) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Building Intelligent .NET Applications: From AI Features to Agentic System","url":"https://daily.dev/posts/building-intelligent-net-applications-from-ai-features-to-agentic-system-rtromjgwu","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/building-intelligent-net-applications-from-ai-features-to-agentic-system-rtromjgwu"},"datePublished":"2026-08-17T16:09:19.010Z","dateModified":"2026-09-14T06:54:40.778Z","description":"A workshop-style talk walks through building AI-powered .NET applications, starting with Microsoft.Extensions.AI (MEAI) as an abstraction layer over LLM...","image":"https://i.ytimg.com/vi/_T5eRmlsfOo/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/_T5eRmlsfOo/sddefault.jpg","isAccessibleForFree":true,"articleSection":".NET Blog","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":".NET Blog","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/f23eb59371bd4f07aeef64c4eef73266","url":"https://daily.dev/sources/dotnet"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/building-intelligent-net-applications-from-ai-features-to-agentic-system-rtromjgwu","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":".net,mcp,rag,vector-search","timeRequired":"PT77M","video":{"@type":"VideoObject","name":"Building Intelligent .NET Applications: From AI Features to Agentic System","description":"A workshop-style talk walks through building AI-powered .NET applications, starting with Microsoft.Extensions.AI (MEAI) as an abstraction layer over LLM...","thumbnailUrl":"https://i.ytimg.com/vi/_T5eRmlsfOo/sddefault.jpg","uploadDate":"2026-08-17T16:09:19.010Z","duration":"PT77M","url":"https://api.daily.dev/r/RTroMjgWU","embedUrl":"https://www.youtube.com/embed/_T5eRmlsfOo"}} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":".NET Blog","item":"https://daily.dev/sources/dotnet"},{"@type":"ListItem","position":3,"name":"Building Intelligent .NET Applications: From AI Features to Agentic System"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/building-intelligent-net-applications-from-ai-features-to-agentic-system-rtromjgwu#faq","mainEntity":[{"@type":"Question","name":"What is Microsoft.Extensions.AI (IChatClient) and how does it help connect to different LLM providers in .NET?","acceptedAnswer":{"@type":"Answer","text":"Microsoft.Extensions.AI provides the IChatClient abstraction, a single .NET interface for talking to any large language model backend instead of learning each provider's own SDK. Implementations exist for OpenAI, Azure OpenAI, Microsoft Foundry, and Ollama, plus community packages on NuGet for providers like Google Gemini. Swapping providers is typically as simple as changing the connection string and model name. Developers comparing LLM providers for .NET apps can track ecosystem updates like this on daily.dev."}},{"@type":"Question","name":"How does structured output work with Microsoft.Extensions.AI in .NET to avoid parsing broken JSON from an LLM?","acceptedAnswer":{"@type":"Answer","text":"Structured output lets you request a response as a specific C# type (for example a record with typed properties), and Microsoft.Extensions.AI sends a JSON schema to the model behind the scenes so the response is enforced to match that shape and deserialized directly into the type. This avoids manual text parsing and issues like models wrapping JSON in markdown or extra commentary, though how well a model honors the schema still depends on the specific model and provider. Teams building typed AI responses in .NET can follow guidance like this on daily.dev while shipping features."}},{"@type":"Question","name":"What is the Microsoft Agent Framework and how does it relate to Semantic Kernel and AutoGen?","acceptedAnswer":{"@type":"Answer","text":"The Microsoft Agent Framework is the unified successor to Semantic Kernel and AutoGen, now generally available (no longer just a release candidate). It wraps an IChatClient together with tools, context, and instructions into a componentized agent, supports multi-agent orchestration with handoffs to pass context between agents, uses the A2A (agent-to-agent) protocol for communication, and can be paired with an agentic UI layer called AGUI. Developers evaluating multi-agent frameworks for .NET can keep up with Agent Framework changes via daily.dev."}}]} ```