Document type
PDF Intelligence at Scale
Pull structured data from millions of PDFs. The AI reads the layout and the meaning, not just the text.
At a glance
OdysseyGPT reads digital and scanned PDFs, including tables and forms. Each value it pulls out cites its page, and unclear values go to a person to check.
Key Takeaways
- Common extraction targets include Text content and paragraphs, Tables and structured data, Headers and footers.
- Reads text straight from digital PDFs with perfect accuracy, and keeps the structure.
- Pull clauses, parties, dates and duties from legal contracts.
Common fields
- Text content and paragraphs
- Tables and structured data
- Headers and footers
- Images and diagrams
- Form fields and annotations
- Metadata and properties
Processing capabilities
- Native PDF Parsing: Reads text straight from digital PDFs with perfect accuracy, and keeps the structure.
- OCR Processing: Reads scanned PDFs with OCR, and cleans up each page first.
- Table Extraction: Finds and pulls out tables, even tables that span pages or have merged cells.
- Form Recognition: Pulls data from PDF forms, both fillable and flattened.
- Layout Analysis: Understands columns, headers, footers and the order to read them in.
- Embedded Content: Pulls out images, attachments and files stored inside a PDF.
Questions answered
What should teams extract from pdf documents?
Start with Text content and paragraphs, Tables and structured data, Headers and footers, Images and diagrams. Add more fields later as your other systems need them.
What are the common risks when automating pdf documents?
Reads text straight from digital PDFs with perfect accuracy, and keeps the structure. Reads scanned PDFs with OCR, and cleans up each page first.
What is the recommended automation flow?
Load the document and extract the fields that matter. Send low-confidence results to a person for review. Then send the checked output to the target workflow or system of record.