A UDC‑based document catalog with LLM‑assisted classification
The project lets you ingest documents, automatically assign Universal Decimal Classification numbers using an LLM, and store each entry as a markdown card. A Flask web UI provides browsing, full‑text search, and editing of cards, while a Python API lets developers query the catalog programmatically. It targets researchers, librarians, and anyone needing a structured, searchable knowledge base without relying on LLMs at query time. By leveraging the UDC hierarchy, it offers a more systematic taxonomy than generic tag‑based systems.
View on GitHub →cadeon/dds-rag