Agentic AI on AWS Hackathon (AAI-HACK-AWS) – Outline

Detailed Course Outline

The challenges are interconnected and built upon one another, guiding participants through the process of creating a complex, multi-agent AI system. The final product will be a prototype of article-writing automation solution.

Challenge 1: The First Agent – Foundations & Accelerated Deployment

Participants will begin by using the Amazon Bedrock AgentCore starter toolkit to rapidly set up their development environment. This toolkit automates the complex process of creating a Docker file, building a container image, and deploying it to the AgentCore Runtime. Teams will then create their first agent, which will serve as the central orchestrator for the future multi-agent system.

Challenge 2: Bridging to the Real World – Tool Integration with Gateway

This challenge focuses on giving the agent the ability to interact with external systems. Instead of building a custom tool server, participants will use the Amazon Bedrock AgentCore Gateway, a fully managed service that transforms existing APIs and AWS Lambda functions into agent-ready tools. Teams will integrate a pre-provided Lambda function, such as one for a web crawler or an article database, and expose it to their agent via the Gateway.

Challenge 3: Orchestration & Intelligence – Multi-Agent Collaboration and RAG

Participants will expand their system by creating specialized sub-agents, such as a "Writer Agent" and an "Art Agent." The challenge culminates in configuring the main orchestrator agent to use Amazon Bedrock's multi-agent collaboration feature to delegate tasks to these specialized agents. Additionally, teams will explore the possibilities of integrating a Knowledge Base to provide the agents with a data source for Retrieval-Augmented Generation (RAG).

Challenge 4: Finalizing the Blueprint – Production Deployment

Description: In this final challenge, teams explore the aspects of running agentic solution in production (including scalability, safety and security) and deploy the entire multi-agent system to the AgentCore Runtime, showcasing how the platform facilitates a seamless transition from a prototype to a secure and scalable production application.