Extract. Ground. Verify.
TrueTrace is Roysa's document understanding engine for extracting structured data from scanned documents while preserving full traceability to the original source.
Upload PDFs or images of your scanned documents
Specify the fields you want to extract
AI extracts values and anchors them to source locations
View annotations and confidence scores
Export structured, auditable results
Process documents with any number of pages seamlessly
Extract exactly what you specify, nothing more
Visual annotations with precise page references
Know exactly how certain each extraction is
Integrate custom validation rules and checks
JSON, CSV, or integrate via API
Every extraction includes complete context for verification
{
"extractions": [
{
"key": "GRANTOR",
"value": "John Smith",
"bounding_box": [120, 340, 280, 365],
"page": 1,
"confidence": 0.98,
"evidence": "...recorded by John Smith..."
},
{
"key": "EFFECTIVE_DATE",
"value": "2024-03-15",
"bounding_box": [450, 520, 560, 545],
"page": 1,
"confidence": 0.95,
"evidence": "...effective March 15, 2024..."
}
]
}Government forms, permit applications, registration documents
Financial documents requiring accurate data extraction
Technical specifications and blueprint details
Compliance and quality assurance documentation
Historical scanned documents and records
Contracts, agreements, and court filings