OCR-assisted document review

Document Assessment

Review selected KYC/AOF, trade, loan, financial-statement and collateral documents in the browser. The workspace extracts selectable text and can use OCR for scanned pages or images; it identifies potential document anomalies for human review. Its current checks look for expected terms and review-language signals across combined extracted text; they do not independently compare invoice, packing-list and transport fields document by document. Manually reconcile parties, goods, quantities, prices, dates and shipment references against the originals.

Document assessment

Switch workspace

Back to TBML Check Back to Transaction Monitoring

Privacy and scope: Files are processed locally in this browser and are not uploaded to a TradeGuard server. Optional OCR loads the open-source Tesseract.js engine and language assets from a CDN; the documents themselves remain in the browser. Use fictional, synthetic or anonymised files only.

STEP 01

Select assessment

Choose the document family, then select up to five local PDF or image files.

Assessment type (required) The selected profile controls the expected document fields and anomaly checks. PDF or image documents (up to 5 files; 10 MB each; 40 MB combined) Selectable-text PDFs are read directly. Scanned PDFs and images are rendered and sent to the browser-local OCR worker. OCR language The Bangla language model is downloaded only when selected and may take longer on the first assessment. I understand that this public demo is for fictional or anonymised documents only and that OCR output requires human verification.

No documents selected.

STEP 02

Document Assessment Report

The report contains metadata and flags, not extracted document text. Email report opens Gmail with From set to [email protected] and To left blank; attach the saved PDF before sending.

▧

Your document assessment will appear here

Select a document family and local files, then run the assessment.

What this module checks

KYC / AOF

Identity, address, occupation, beneficial ownership, source-of-funds and signature completeness signals.

Trade documents

Invoice, packing list, bill of lading, parties, HS code, quantity, price, route and document-consistency signals.

Credit documents

Loan applications, financial statements and collateral files for completeness, consistency and missing-control signals.

OCR limitation: OCR quality depends on resolution, language, layout and document condition. The output is a screening aid only; it must not replace review of the original document or an institutionally governed credit, KYC or AML process.