AWS Bedrock AgentCore Essentials Training in Marina Bay, Singapore

This course introduces Amazon Bedrock AgentCore, AWS’s managed framework for building, deploying, securing, and monitoring AI agents. Participants will explore AgentCore architecture, memory, identity, tool integration, observability, and best practices for production-ready deployments.

• Introduction to AgentCore

·        Purpose, use cases, and ecosystem role

·        Components: Runtime, Memory, Identity, Gateway, Browser, Code Interpreter, Observability

• Environment Setup
• SDKs, IAM roles, cluster requirements
• First deployment with AgentCore Runtime
• Gateway & Tool Integration
• Using Gateway to connect APIs and services
• Configuring agent toolchains
• Memory in AgentCore
• Short-term vs long-term memory
• Context management strategies
• Identity & Security
• IdentityClient, workload identity, token vault
• Permission boundaries and access policies
• Observability
• Metrics, tracing, latency tracking, error handling

• Advanced Tools in AgentCore
• Code Interpreter: secure sandbox execution
• Browser Tool: web automation and scraping use cases
• Scaling & Deployment Best Practices
• Session isolation, error handling, low latency design
• Compatibility with different model frameworks
• Governance, Compliance, and Security Best Practices
• Identity federation, auditing, and compliance controls
• Agent safety and fallback strategies
• Review & Best Practices
• Troubleshooting strategies, monitoring patterns, future roadmap

Hands-on Exercises

Summary and Conclusion