Loading…
AzureDocumentIntelligenceChunker is a lightweight Python library that provides a production-grade solution for chunking documents in Retrieval-Augmented Generation (RAG) workflows. It leverages Azure AI Document Intelligence to preserve document hierarchy, including page numbers and bounding boxes, enabling accurate highlighting and easy PDF viewer integration. It runs with native Python packages (no Langchain dependency), mirrors Azure API markdown output, and automatically prefixes chunks with headers to enhance recall and retrieval relevance.
AI named AzureDocumentIntelligenceChunker in December 2025.