Read
OCR for scanned documents
Make scanned PDF text available for extraction and source review.
Explore capabilityThe document processing platform
From a document
to data you can use.
Read PDFs and scans. Extract fields and tables. Check the source. Deliver structured data to the next step in your workflow.
Explore the capabilities available today and the roadmap for AI extraction, organization knowledge and tested document workflows.
- 01Read
- 02Extract
- 03Review
- 04Find
- 05Deliver
Use the product today
A shared foundation for document work.
Every capability page describes supported inputs, outputs and limits. Start with a representative document and inspect the result.
Extract
Structured document extraction
Turn supported document layouts into fields and rows you can check.
Explore capabilityExtract
PDF table extraction
Move table data into a spreadsheet without losing its source.
Explore capabilityReview
Source-linked document review
Make every accepted value a decision you can trace.
Explore capabilityReview
Document data validation
Put exceptions in front of the person who can resolve them.
Explore capabilityOperate
Batch document processing
Repeat a document workflow with the same setup and output format.
Explore capabilityDeliver
Excel, CSV and JSON exports
Hand off useful data with enough context to verify it.
Explore capabilityFind
Document search and retrieval
Get back to the evidence inside your document collection.
Explore capabilityDeliver
Document APIs and workflow integrations
Connect document processing to the next step in your workflow.
Explore capabilityAutomate
Durable document workflows
Turn a repeated document procedure into a workflow you can inspect, test and run again.
Explore capabilityFind
Organization knowledge
Build reusable organization records with inspectable evidence and review history.
Explore capabilityImprove
Verified workflow optimization
Replace repeated extraction decisions with rules you have tested on separate examples.
Explore capabilityImprove
Evaluation and drift monitoring
See what an extraction rule gets right, misses or cannot support before promoting it.
Explore capabilityMatch
Statement reconciliation
Find matched records, duplicate references and unexplained differences in a defined statement period.
Explore capabilityOperate
Document access and governance
Keep document access and data lifecycle understandable.
Explore capabilityProduct roadmap
From document processing to dependable workflows.
These capabilities are planned and are not included in the current trial. The direction: use AI to help define a process, build organization knowledge, and evaluate rules, scripts and model steps before repeating the work.
Extract
AI schema extraction
Define the data you need and receive structured values with citations.
Explore capabilityFind
GraphRAG and connected document search
Follow connections across documents and return to the evidence for each one.
Explore capabilityOrganize
Document classification and splitting
Give each document in an incoming packet the right processing path.
Explore capabilityOperate
Model routing and budgets
Use the least costly evaluated path that meets the task’s requirements.
Explore capabilityPlanned routing and optimization will compare quality, review effort and total cost, with monitoring when inputs change. Release dates and performance claims depend on working implementations and evaluation. Multilingual OCR, image uploads, hybrid search, review-queue assignment and destination connectors also remain planned.
Built around your work
One platform. Specific workflows.
Bookkeeping remains a core use case, alongside supplier data and recurring document operations.
Start with the source.
Inspect the synthetic sample, then try your own supported PDF. Eligible workspaces receive 10 pages for 7 days.