This AgentCore course uses Python, PyCharm, and Python Virtual environments to cover every aspect of Amazon Bedrock AgentCore in depth. You’ll learn how to run agents locally while connecting to LLMs in the cloud, and how to deploy agents and multi-agent systems to AgentCore on AWS. The architecures are discussed for different frameworks and deployments,
The value of AgentCore lies in its runtime — giving you observability, the ability to evaluate your agents, reusable prompts, and the tools to define multi-agent systems that communicate with each other. Each demo in this course is built around learning Python and the core AgentCore APIs hands-on.
This is a thorough, architecture-focused development course, with 90% of the content dedicated to hands-on coding. You’ll run notebooks, execute different cells, and observe outputs, while also learning how Docker and CloudFormation are used to deploy to AgentCore.
Once deployed, we’ll explore the AWS console together, diving into observability features — examining traces and spans to identify performance bottlenecks and learn how to optimize your agent-based systems. While running these examples we use Claude Code to fix issues and successfully run the Python Notebook when necessary.
