opendataloader-pdf

The ultimate open-source PDF remediation engine, combining AI-automated structural tagging with Human-in-the-Loop interface for perfect PDF/UA validation. Shift PDF compliance left. The pdf-a11y is a complete toolkit—CLI, REST API, and Web UI—that automatically fixes document tags, generat

Django API that checks uploaded PDFs against accessibility standards (WCAG 2.1 AA / PDF/UA-1 via VeraPDF + a custom Claude script) and automatically remediates the ones that fail - OCR, tagging, metadata fixes, alt text, and link repair, before re-checking them and returning via webhook. Developed by Van Buren County DID.

Production-grade PDF language detection using fastText lid.176.ftz. Detects language per page, groups consecutive pages into segments, and splits PDFs by language — fully typed, thread-safe, and batch-ready.