How Observability Makes Better AI and Human Investigators

Agents are only as good as their context

Join Honeycomb and AWS for a live, virtual workshop where you’ll learn what happens when agents can’t access good telemetry data.

We’ll provision an AWS environment for you. You’ll keep access for 48 hours after the session so you can finish, rerun, or break things on your own time. The only thing you need to worry about is creating a free Honeycomb account ahead of the workshop.

During the workshop, you’ll:

• Deploy a real microservices app on Amazon EKS and send its telemetry to Honeycomb with one exporter change
• Investigate a real latency spike using traces, BubbleUp, and Canvas
• Watch an AWS DevOps Agent investigate the same issue twice: blind, then with Honeycomb connected through an MCP to compare its findings
• Build and instrument your own AI agent in about 60 lines of code and watch its full Agent Timeline appear: reasoning, tool calls, and token usage

You leave with a clear picture of how OpenTelemetry, Honeycomb, and AWS support both human and agent investigation.

Prerequisites

All workshop activities use Honeycomb features available in the free tier, so you can continue exploring the workflow after the session. Be sure to sign up for a free account!