Bring AI agents into production in minutes with Amazon Bedrock AgentCore
This repository contains hands-on labs demonstrating the capabilities of Amazon Bedrock AgentCore, an agentic platform to build, deploy and operate agents securely at scale - using any framework and model.
Amazon Bedrock AgentCore enables developers to accelerate AI agents into production with enterprise-grade scale, reliability, and security. AgentCore provides composable services that work with popular open-source frameworks and any model, eliminating the choice between open-source flexibility and enterprise requirements.
| Service | Purpose | Key Features |
|---|---|---|
| AgentCore Runtime⭐ | Serverless execution | Auto-scaling, session management, container orchestration |
| AgentCore Identity | Credential management | API keys, OAuth tokens, secure vault |
| AgentCore Memory⭐ | State persistence | Short-term memory, long-term storage |
| AgentCore Gateway | Connects agent to tools and data | Tool discovery, service integration |
| AgentCore Code Interpreter | Code execution | Secure sandbox, data analysis |
| AgentCore Browser | Web interaction | Cloud browser, auto-scaling |
| AgentCore Observability | Monitoring | Tracing, dashboards, debugging |
| AgentCore Policy | Security boundaries | Deterministic control, Cedar policies, natural language authoring |
| AgentCore Evaluations | Performance assessment | Automated testing, LLM-as-a-Judge, quality metrics |
Before starting any lab, ensure you have:
- AWS Account with appropriate permissions
- Python 3.10+ installed
- AWS CLI configured
- Basic understanding of AI agents and AWS services
For faster Python dependency management, consider using uv instead of traditional pip and venv:
# Install dependencies with uv (faster alternative to pip)
uv pip install -r requirements.txt
# Or initialize projects with uv
uv init my-agent-projectThis is optional - all labs work with standard pip commands as documented.
| 📓 Services | 🎯 Focus & Key Learning | ⏱️ Time | 📊 Level |
|---|---|---|---|
| 01 - Amazon Bedrock AgentCore Runtime | Serverless AI agent deployment with auto-scaling, session management, and built-in security | 10 min |  |
| 02 - Amazon Bedrock AgentCore Memory | Context-aware memory for conversation context and cross-session knowledge retention | 10 min | |
| 📓 Services | 🎯 Focus & Key Learning | 🖼️ Diagram |
|---|---|---|
| Amazon Bedrock AgentCore Runtime | Focus: Serverless AI Agent Deployment Deploy production-ready AI agents with just 2 commands using AgentCore Runtime. This lab demonstrates: • Serverless agent deployment with auto-scaling • Session management and isolation • Built-in security and authentication • Integration with Strands Agents framework Key Learning: Transform prototype agents into production-ready services in minutes, not weeks. |
 |
| Amazon Bedrock AgentCore Memory | Focus: Intelligent Memory Capabilities Add context-aware memory to AI agents using AgentCore Memory. This lab covers: • Short-term memory for conversation context • Long-term memory for user preferences • Cross-session knowledge retention • Personalized agent experiences Key Learning: Build agents that remember and learn from interactions to provide more intelligent responses. |
 |
Each lab includes:
- Prerequisites: Required setup and dependencies
- Step-by-step deployment: Automated infrastructure setup
- Code explanations: Detailed implementation walkthrough
- Cleanup instructions: Resource removal
Ready to deploy production AI agents? Start with 01-agentcore-runtime to learn the fundamentals of AgentCore Runtime.
- What is Amazon Bedrock AgentCore?
- AgentCore Runtime How It Works
- AgentCore Memory Guide
- AgentCore Gateway Documentation
- Programmatic Agent Invocation
See CONTRIBUTING for more information.
This library is licensed under the MIT-0 License. See the LICENSE file.
