.NET AI Community Standup: Build agentic experiences with the Blazor AI Components
Bruno Capuano and Daniel Roth introduce the Blazor AI Components and show how they enable agentic experiences in .NET apps beyond basic chat, including structured streaming responses, tool invocation, approval gates, and multi-step workflows coordinated through shared state.
Overview
Modern AI applications can be more than chat. This community standup focuses on how Blazor AI Components can be used to build agentic experiences in .NET applications, where agents participate in user workflows and the UI stays responsive and user-controlled.
Key capabilities discussed:
- Agents in user workflows
- Agents participate in multi-step tasks instead of only answering prompts.
- Streaming structured responses
- UI patterns for rendering incremental output while preserving structure.
- Tool invocation (tool calling)
- Agents can invoke tools as part of completing tasks.
- Approvals / human-in-the-loop
- Agents can request explicit user approval before taking certain actions.
- Collaboration through shared state
- Coordinating interactions and progress across steps using shared state.
- Blazor component model for AI interactions
- Using components to render agent interactions and build responsive, user-controlled experiences.