Features:

• Managed RAG: Build Retrieval-Augmented Generation (RAG) applications without managing the underlying retrieval infrastructure, vector databases, or indexing pipelines.
• Data ingestion & processing: Ingest data from sources such as Amazon S3 and supported enterprise data sources, with built-in document parsing, chunking, embedding, and indexing capabilities.
• Semantic & vector search: Retrieve relevant information from enterprise data using semantic and vector-based search capabilities.
• Metadata filtering: Filter retrieval results using document metadata to provide more precise and context-aware results.
• Integration with foundation models: Connect retrieved knowledge with Amazon Bedrock foundation models to build context-aware AI applications and chatbots.