START WITH YOUR TASK
The work comes first.
Use these categories as a map, not a mandatory pipeline. A tool can belong to more than one task.
Agentic Document AI ↗
Coordinate multi-step document workflows, inspect evidence, and revise results.
3 tools · Planning, extraction, verificationTASK 02Document Parsing & OCR ↗
Recover text and document content from digital files and scanned pages.
6 tools · Text layers, OCR, scanned documentsTASK 03Layout & Table Reconstruction ↗
Recover reading order, table cells, and page structure before downstream extraction.
6 tools · Reading order, tables, layoutTASK 04Field Extraction ↗
Extract typed fields against a defined schema and inspect the evidence behind each value.
7 tools · Schemas, fields, confidenceTASK 05Classification & Splitting ↗
Identify document types and page boundaries inside mixed document packages.
3 tools · Classification, page groups, routingTASK 06Format Conversion & Document Databases ↗
Convert files into usable formats or build structured records and relational databases.
5 tools · Markdown, records, databasesTASK 07Ontology & Semantic Modeling ↗
Define concepts, relationships, mappings, and constraints for data and agents.
5 tools · Ontology construction, evolution, alignmentTASK 08Multimodal Retrieval & RAG ↗
Retrieve page-level visual evidence and ground answers in relevant document content.
4 tools · Page retrieval, evidence, answersTASK 09Visual Document Understanding ↗
Use visual models to interpret pages, charts, and document relationships.
1 tools · Visual reasoning, OCR-free understanding