Building Advanced Agentic Systems on AWS
This 1-day advanced course shows how to take multi-agent AI systems beyond proof-of-concept and into production using Amazon Bedrock AgentCore.
The Building Advanced Agentic Systems on AWS course looks at what it takes to make agentic AI scalable, secure, and observable. Participants start with the limits of single agents and the benefits of multi-agent architectures, then explore communication patterns, memory and state management, and the agent-as-tool pattern for production deployments.
Along the way the course covers context engineering, where context is treated as a finite resource, including optimization and compression techniques, efficient tool design, and keeping resource usage and costs under control. Security and compliance get their own module, with identity and access management, policy-based access control using the AgentCore Policy Engine, VPC integration, and audit trails. Participants also learn about distributed tracing, monitoring, agent evaluation, and integration with enterprise observability systems. The course closes by applying the Well-Architected approach to agentic systems and assessing production readiness.
The day combines presentations, a hands-on lab, and group exercises, including instructor demonstrations in which financial agents are built and deployed with Strands Agents and Amazon Bedrock AgentCore.
The course is aimed at software developers, technical professionals exploring AI capabilities, and development teams building advanced agentic AI solutions. It is recommended that attendees have completed Agentic AI Foundations and Building Agentic AI with Amazon Bedrock AgentCore.