Document Intelligence
Connect sources, extract and review information, govern records and deliver usable outputs
Get started with Document Intelligence
Follow a business record from its original source through extraction, review and controlled delivery.
Organise the document library
Find files efficiently and manage metadata, versions, folders, access and retention.
Connect sources and prepare documents
Understand source admission, malware clearance, supported preparation and source lineage.
Choose native text extraction and OCR
Configure OCR behaviour, Tesseract, AWS Textract and review requirements without confusing OCR with approval.
Configure classification and extraction schemas
Create a draft class, define fields, test representative records and publish a controlled version.
Run and monitor bulk processing
Create a bounded batch, track its documents and use job insights without confusing queue completion with delivery.
Review extracted information and complete a document
Compare source evidence, correct fields, resolve validation and follow indexing and QC gates.
Redact sensitive information and govern retention
Apply tested redaction policies, review generated copies and preserve the original and legal obligations.
Create outputs and deliver them safely
Preview datasets and source-file options, verify destinations and interpret delivery receipts.
Understand analytics, usage and Assurance
Read operational measures and the Production, Codebase and Pipeline views with the correct scope.
Resolve common Document Intelligence blockers
Find the stage that needs attention and take a safe next action.
Configure extraction and verify indexed values
Define the values you need, test sample documents and check the current governed index separately from extraction approval.
Connect and govern data sources
Understand source boundaries, provider settings and the difference between connecting a repository and processing its records.
Test a classification schema before publishing
Build a small repeatable test set and check recognition, fields and review rules against expected outcomes.
Check whether a document is actually indexed
Compare reviewed values with Current index and the recorded indexing result.
Work a sampled quality-control review
Inspect the fixed sample, record a supported outcome and resolve failures before release.
Preview an output package before transfer
Check data, files and destination against the intended receiving system before releasing a transfer.