A mini course on building AI agents with the Strands Agents framework, from beginner to advanced levels.
WARNING: A tiny Ads here for building your Streamlit App Demonstrating Strands Agents.
This course takes you through a complete journey of learning how to build, customize, deploy, and orchestrate AI agents using the Strands Agents framework. All examples use the Amazon Bedrock Nova Lite model (us.amazon.nova-lite-v1:0).
Tips: You can also find more information and ask questions on DeepWiki page.
- Python 3.12 or higher
- Basic understanding of Python programming
- Jupyter Notebook environment
- AWS account with access to Amazon Bedrock (for Nova Pro model)
pip install -U strands-agents strands-agents-tools- Understanding the Strands Agents framework
- Creating your first agent
- Running basic agent interactions
- Setting up your development environment
- Exploring built-in tools
- Understanding tool interfaces
- Using calculator, python_repl, and other core tools
- Basic tool orchestration
- Creating custom tool functions
- Tool decorators and annotations
- Tool error handling
- Advanced tool patterns
- Introduction to the Model Context Protocol
- Using MCP-based tools
- Creating your own MCP servers
- Preparing agents for production
- AWS Lambda deployment
- *API Gateway integration
- Agent as Tools pattern
- Swarm
- Graph
- Workflow
- Logging full lifecycle operations
- Monitoring performance and latency
- Tracing spans of actions
- Start with Chapter 1 and work your way through each chapter sequentially
- Run the code examples in each notebook to gain hands-on experience
- Complete the exercises at the end of each chapter to reinforce your learning
- Experiment with your own variations of the examples
The later chapters cover advanced topics including:
- Production deployment on AWS Lambda
- Multi-agent orchestration
- Best practices for complex agent systems
- All examples use the Nova Lite model from Amazon Bedrock
- Make sure to set up your AWS credentials appropriately when working with the examples
- Each notebook is self-contained and can be run independently, though they build on concepts from earlier chapters
- Other examples could be found here.
Happy learning! 🎉