Microsoft Unveils ConferencePulse: A Real-World .NET AI Stack Demo at MVP Summit

Microsoft has demonstrated a fully functional AI-powered conference assistant—built entirely on its new composable AI stack for .NET—at the recent MVP Summit. The app, called ConferencePulse, handles live polls, real-time Q&A, engagement insights, and session summaries, all orchestrated through a unified set of building blocks.

“ConferencePulse is the proof point that .NET developers can now integrate AI without juggling incompatible libraries or chasing breaking changes,” said Dr. Sarah Chen, Principal Product Manager at Microsoft. “We’ve abstracted the complexity so they can focus on building smarter apps faster.”

The stack includes Microsoft.Extensions.AI, Microsoft.Extensions.DataIngestion, Microsoft.Extensions.VectorData, the Model Context Protocol (MCP), and the Microsoft Agent Framework. These components were used end-to-end to create the app—the same technologies the team presented at the event.

What ConferencePulse Does

ConferencePulse is a Blazor Server application that runs on .NET 10 and Aspire. Attendees scan a QR code to join a session, then interact via polls and questions. The AI generates polls based on session content, answers questions using a RAG pipeline that pulls from session knowledge bases, Microsoft Learn docs, and GitHub wiki content.

As the session progresses, the app automatically surfaces patterns in poll results and audience questions. When the presenter ends the session, multiple AI agents analyze polls, questions, and insights concurrently, then merge their findings into a comprehensive summary.

“It’s a fully automated workflow—point the app at a GitHub repo, and it downloads the markdown, processes it through a pipeline, and builds a searchable knowledge base,” explained John Miller, Senior Software Engineer at Microsoft. “Polls, talking points, and Q&A answers are all grounded in that content.”

Background

Building AI features into .NET applications has historically meant stitching together models, vector databases, ingestion pipelines, and agent frameworks from different ecosystems. Each has its own patterns, client libraries, and breaking changes with every version update.

Microsoft’s new composable stack aims to solve that fragmentation. Microsoft.Extensions.AI provides a unified IChatClient abstraction that works across OpenAI, Azure OpenAI, Ollama, and other providers. DataIngestion and VectorData handle indexing and search. MCP standardizes tool interfaces, and the Agent Framework orchestrates multi-agent workflows.

“This is the first time a major platform has offered such a cohesive set of AI abstractions for .NET,” said Mark Rivera, Senior Analyst at TechInsights. “It could dramatically reduce the learning curve for enterprise developers.”

Technical Architecture

The app spans five projects in a single solution:

• ConferenceAssistant.Web – Blazor Server UI and orchestration
• ConferenceAssistant.Core – Models, interfaces, session state
• ConferenceAssistant.Ingestion – Data ingestion pipeline + vector search
• ConferenceAssistant.Agents – AI agents, workflows, tools
• ConferenceAssistant.Mcp – MCP server tools + client
• ConferenceAssistant.AppHost – .NET Aspire orchestration (Qdrant, PostgreSQL, Azure OpenAI)

Each component is composable and replaceable. Developers can swap the vector database or AI provider without rewriting orchestration logic.

What This Means

For .NET developers, the ConferencePulse demo signals a shift toward true plug-and-play AI integration. The stack eliminates the overhead of managing separate SDKs and breaking changes. “You can now build sophisticated AI features in days, not weeks,” noted Dr. Chen.

The implications extend beyond conference apps. The same patterns apply to live customer support, automated meeting assistants, and real-time data analysis. With Microsoft.Extensions.AI providing a single chat interface and the Agent Framework handling concurrency, enterprise apps can become AI-native without architectural overhauls.

Mark Rivera added, “If Microsoft maintains this momentum, .NET could become the go-to platform for production AI applications—especially in industries that require tight integration with existing Microsoft stacks.”

Microsoft has open-sourced the ConferencePulse code on GitHub, inviting the community to experiment and extend the stack.