This article shows .NET developers how to get started building and debugging agent-based applications with Microsoft Agent Framework and DevUI. Using the aiagent-webapi template, readers create a multi-agent workflow that includes Writer, Editor, and Publisher agents while learning how Microsoft Agent Framework integrates with familiar ASP.NET Core patterns such as dependency injection and hosted services.
The article explores how DevUI shortens the agent development loop by providing a visual interface for running workflows, inspecting agent interactions, and debugging execution locally. As workflows become more complex, the article highlights the growing observability gap between local debugging and production AI systems, where teams need visibility into behavior across sessions, latency, cost, failures, and evaluations.
To bridge that gap, the article introduces Progress AI Observability Platform and demonstrates how to instrument applications using OpenTelemetry and the .NET Activity pipeline. Readers learn how to capture traces, monitor workflows across sessions, analyze costs, and drill into telemetry data for deeper debugging and operational insight.
By combining DevUI with observability tooling, the article demonstrates a more complete agent development loop that evolves from local experimentation into production-ready diagnostics, tracing, evaluation, and monitoring for enterprise AI applications.
While there are numerous ways to deploy an app to Cloud Run, see how to containerize and deploy a NestJS API. We will use a few products on GCP, such as Buildpacks and Artifact Registry, to build and deploy our image, and then finally deploy it to Cloud Run.
Learn how to measure and optimize AI token spend before billing surprises hit. Discover why production AI costs diverge from estimates and how trace-level observability helps teams control LLM spending.
The RAG resources you create to use with your LLMs are strategic resources, just like your organization’s databases. Telerik Agent Tools let you create the custom tools for managing those resources.
We’ll build a multi-tenant task management API where customer data is isolated automatically at the database level. We’ll use Postgres Row-Level Security to enforce tenant isolation, with NestJS as our application framework and TypeORM to interact with the database.