Demystifying Microsoft Agent Framework Middleware
Featuring: Daniel Costea, Katie Savage, Scott Addie, Frank Boucher #dotnet #dotnetdeveloper
In this session we'll walk through how Daniel Costea uses ChatClient and Agent middleware to turn plain .NET AI agents into production‑ready systems with observability, safety, and cost control. Starting from an agent that talks directly to an LLM, he walks through real‑world failure modes (cost spikes, privacy leaks, unsafe tool calls) and then layers in SharedFunction, Response, and FunctionCalling middleware to fix them. Along the way, he connects each concept to the familiar ASP.NET Core middleware pattern so .NET developers can reuse an existing mental model for building robust AI agents.

Featuring: Daniel Costea, Katie Savage, Scott Addie, Frank Boucher #dotnet #dotnetdeveloper
In this session we'll walk through how Daniel Costea uses ChatClient and Agent middleware to turn plain .NET AI agents into production‑ready systems with observability, safety, and cost control. Starting from an agent that talks directly to an LLM, he walks through real‑world failure modes (cost spikes, privacy leaks, unsafe tool calls) and then layers in SharedFunction, Response, and FunctionCalling middleware to fix them. Along the way, he connects each concept to the familiar ASP.NET Core middleware pattern so .NET developers can reuse an existing mental model for building robust AI agents.

Featuring: Leslie Richardson, Simona Liao #visualstudio
Simona shows us how to build and use agent skills, which are lightweight packages with specialized knowledge an agent can discover and use while solving a task.

Featuring: Shay Rojansky, Jiri Cincura #ef #efcore #data #temporal #sqlserver #postgres
In this session we'll take a deep look at temporal tables and constraints in database. The recently-released PostgreSQL 18 contains new temporal constraint functionality that superficially looks similar to SQL Server's temporal tables, but the two are in reality quite different. We'll dive into the details, compare the two and discuss two completely different concepts of what it means to store temporal data in databases.




