Webinar

How to build AI agents with MCP and other protocols

Event Overview
Agent protocols can turbocharge your AI agent development by standardizing communication and integration of your agent with  tools, data sources, and other agents. In this webinar, we’ll explore how to build reliable agentic systems using protocols such as Model Context Protocol (MCP) and Agent2Agent (A2A) with Weights & Biases.
What to expect
  • A step-by-step walkthrough to evaluate and improve AI agents for production-level reliability using W&B Weave’s built-in integrations with protocols like MCP and A2A.
  • A live demo showcasing observability for agentic systems built with popular frameworks, leveraging our pre-built integrations and OpenTelemetry (OTel) support.
What you'll learn
  • How to build AI applications using agent protocols (MCP and A2A) and frameworks (e.g., OpenAI Agents SDK)
  • How to utilize W&B Weave’s out-of-the-box support for MCP and A2A to log agent traces for evaluation, debugging, and monitoring
  • How Weave collects OpenTelemetry traces from any framework compliant with the standard—no extra instrumentation required
Who should attend
  • AI developers
  • ML engineers

Featured
speakers

Anish
Shah
Machine Learning Engineer
Weights & Biases