Build Production-Ready AI Agents on AWS with Amazon Bedrock AgentCore

In this hands-on workshop, you'll go from zero to a fully deployed multi-agent system on AWS — no prior AgentCore experience needed.

We'll build a real multi-agent research assistant, step by step:

- Deploy your first AI agent using the Strands framework and BedrockAgentCoreApp — running in the cloud in minutes
- Orchestrate multiple agents with the A2A (Agent-to-Agent) protocol, a cloud-agnostic standard that lets agents discover and call each other
securely via IAM
- Connect external tools — wire in PubMed via MCP Gateway, deploy a citation manager as a Lambda function, and expose your own MCP server

By the end you'll have a working orchestrator agent that routes queries to a search specialist, calls real APIs, and manages citations — all
running on AgentCore Runtime with proper IAM permissions and CloudWatch observability.

What to bring:
- Laptop with Python 3.11+ and AWS CLI installed
- An AWS account (free tier works for most of the workshop)
Skill level: Intermediate — some Python and basic AWS familiarity helpful. No prior agent development experience required.