Amazon Kendra launches Retrieval API
Amazon Kendra is an intelligent search service powered by machine learning, enabling organizations to provide relevant information to customers and employees, when they need it. Starting today, AWS customers can use Amazon Kendra as a retriever to build retrieval augmented generation (RAG) systems.
Amazon Kendra is an intelligent search service powered by machine learning, enabling organizations to provide relevant information to customers and employees, when they need it. Starting today, AWS customers can use Amazon Kendra as a retriever to build retrieval augmented generation (RAG) systems.
Recent advancements in large language models (LLMs) has led to growing interest in using RAG to build question-answering capabilities for enterprise content. The retrieval API in Kendra enables customers to combine Kendra with any LLM to build a RAG system. For a given query, the retrieval API finds semantically relevant passages from indexed content which can be sent to an LLM to generate comprehensive answers. Customers can leverage existing features of Kendra like custom synonyms, relevance tuning, custom document enrichment and filtering to improve the relevance of retrieved content. In addition, Kendra has native connectors for 30+ data sources like S3, SharePoint, Confluence, Google Drive, Box and more which makes it easier to ingest enterprise content into the index.
The retrieval API is available in all AWS regions where Amazon Kendra is available including AWS GovCloud (US) Regions. To learn more about the feature, read this blog and visit the documentation page.