Semantic Prism
Make mixed documents usable.
Collect mixed bundles, classify their contents, split them into documents, then capture and validate data before delivery.
- Collect A single PDF contains an invoice, a delivery note and a credit note.
- Classify Identify the invoice, delivery note and credit note within the PDF and determine their page ranges.
- Split Separate the page ranges into three documents and select the matching Document Pack for each one.
- Capture Extract the fields specified in each Document Pack, including references and amounts.
- Validate A separate stage checks captured values against the source documents before delivery.
- Deliver The source PDFs and structured data are delivered to their configured destinations.
The source PDFs and structured data are delivered to their configured destinations.
What it does
Capture the work.
- Versioned Document Packs
- Define document types, fields, instructions and validation outcomes for repeatable capture.
- Page-level classification
- Identify and split mixed source files into the documents you need.
- Validation and recovery
- Check results, track processing and recover work when a stage needs attention.
- Independent delivery
- Send the original material and structured data to the destinations that need each.
AI in this product
Use separate AI stages to read and check documents.
Classification identifies document types and page ranges in a mixed bundle. Prism splits the bundle into documents, then captures fields for each one using its Document Pack. A separate validation stage checks the captured values before delivery.
How you stay in control. Document Packs define the schema, instructions and validation outcomes. Operators can see what happened and recover items that need attention.
Work with Intellim
Which process would you change?
Bring us the inputs, the decisions people make and the systems the work has to reach.
Discuss your workflow