What changed
AWS and Spring AI have introduced capabilities for Java developers to build autonomous AI agents using Amazon Bedrock. These tools address challenges in scaling, governance, security, and operational concerns when transitioning AI agents from prototype to production.
What this means in practice
Java teams can now build AI agents that plan, call tools and APIs, and maintain context across multi-turn conversations. The solution leverages Spring AI's familiar patterns like auto-configuration, dependency injection, and composable advisors, allowing developers to use the same idioms they apply for REST APIs, security, and data access.
The solution includes two deployable systems: a single-agent travel assistant and a multi-agent intelligent document processing (IDP) system. The travel assistant demonstrates tool calling and memory, while the IDP system showcases how specialist agents, a rules engine, and alerting work together. Both systems share a common architecture using Spring Boot, Spring AI ChatClient, and Amazon Bedrock Converse API.
Limitations and availability
The solution requires specific tools and environments, including Java JDK 21, Spring Boot 4.1, Maven 3.9+, Node.js 18+, Terraform 1.5+, Docker with Buildx, and the AWS CLI. The travel assistant is recommended as the starting point, while the IDP system demonstrates production patterns at scale. The tools are available through AWS and Spring AI, with no specific mention of licensing conditions or download availability in the snapshot.
