agent-framework-dotnet

To build AI agents in .NET with multi-turn conversations, use the Agent Framework to configure session-based state management via AgentSession, allowing agents to retain context and coordinate complex conversational workflows graph-based nodes.

Exposing C# methods as function tools for AI agents is done using AIFunctionFactory within the Agent Framework, enabling native method exposure for OpenAI or Azure OpenAI models to invoke directly during chat completions and workflow execution.

Yes, you can integrate Azure OpenAI with .NET workflows for scalable deployments by configuring the Agent Framework's provider settings, enabling robust multi-agent graph execution and tool invocation connected to your Azure OpenAI endpoints.

Adding middleware for logging and error handling in .NET AI agent workflows involves applying the Agent Framework's middleware support to intercept pipeline executions, injecting cross-cutting concerns alongside observability tracking through OpenTelemetry integration.

Yes, the .NET Agent Framework supports streaming responses from OpenAI models, allowing real-time token delivery in multi-turn conversations while maintaining session state and graph-based workflow coordination across distributed agent components.