Building Agentic AI with Amazon Bedrock AgentCore
This 1-day course helps developers turn proof-of-concept agents into production-ready agentic AI solutions with Amazon Bedrock AgentCore.
The Building Agentic AI with Amazon Bedrock AgentCore course starts with the characteristics of agentic AI, how it differs from traditional AI systems, and the core building blocks of an agent. From there, participants work through the AgentCore services that support secure, scalable enterprise deployments.
The course covers deploying agents built with supported frameworks on AgentCore Runtime, including serverless execution with session isolation, and securing them with AgentCore Identity and AgentCore Policy. It then moves on to tool integration, from built-in tools to the Model Context Protocol (MCP), and to AgentCore Gateway for secure and authorized tool access. Further topics are agentic memory with AgentCore Memory, production monitoring with AgentCore Observability and Amazon CloudWatch, and verifying agent performance with AgentCore Evaluations.
The day includes presentations, group exercises, and a hands-on lab on enhancing and scaling agents with Amazon Bedrock AgentCore. It closes with deployment strategies, production readiness, and continuous improvement.
The course is intended for software developers with intermediate knowledge who want to build agentic systems, technical professionals exploring AI capabilities, and development teams building agentic AI solutions. It is recommended that attendees have completed Agentic AI Foundations.