Modular RAG Architectures with Java and Spring AI

Learn to design and implement modular Retrieval-Augmented Generation (RAG) architectures using Java and Spring AI in this comprehensive conference talk from Spring I/O 2025. Discover how to enhance Large Language Models by integrating them with targeted, domain-specific data through hands-on examples and practical guidance for building production-ready RAG workflows. Master the five key stages of RAG implementation: data indexing techniques for structuring, chunking, and indexing documents to improve retrieval efficiency and accuracy; query analysis methods for transforming and enhancing queries in the pre-retrieval phase; retrieval strategies for sourcing data from multiple repositories using agent-based routing and combining results; augmentation approaches for enhancing user prompts with retrieved content including contextualization and citation techniques; and evaluation metrics for assessing document relevance and model output accuracy with iterative feedback loops. Explore important considerations for observability and developer experience while building scalable RAG systems that support both current and future use cases, with access to accompanying slides and code repositories for practical implementation.