Fraud edits the page.
Vrfy reads the file.
Inspect payslips, bank statements and invoices at the file level for signs of tampering and manipulation, in seconds. Clear documents move on; flagged ones get your attention.
You can't see what changed.
The strongest signs of tampering rarely render on the page. Your team is reviewing the pixels. The evidence is in the file.
- Incremental saves
- PDFs can carry their own revision history. A "scanned original" with multiple saved edits is telling you something.
- Font objects
- Text inserted after creation rarely matches the fonts around it. Invisible to the eye, obvious to the file.
- Producer chains
- Files can record the software that made them. A payslip claiming Sage but stamped by a browser print function has already confessed.
Three steps to an answer
No integrations to build. No models to train. Results in seconds.
- 01
Check it where it sits
Drag a PDF or image into the dashboard, or run Vrfy on a document already open in your browser.
- 02
Analyse
The engine runs structural, metadata and arithmetic checks across the document.
- 03
Answer
Get a Clear, Review, High Concern or Inconclusive assessment, with every finding explained and a record of exactly what was checked.
Live assessment preview
No significant anomalies identified within the checks performed.
Traceable rules, not a black box
Every assessment is backed by a rule ID, the observed evidence and a plain-language explanation of why it matters. Same document, same result, every time: auditable flags your team can defend.
What Vrfy inspects
Every document runs the same deterministic battery — structure, metadata and arithmetic — with each finding traced to a rule and the whole run recorded.
Metadata integrity
Creation dates, edit history and software fingerprints inspected for signs of tampering.
Arithmetic checks
Gross pay, deductions, subtotals and VAT recomputed and reconciled to the stated figure.
Balance reconciliation
Every transaction on a bank statement walked against its running balance, row by row.
PDFs and images
PDFs, scans and photographed documents run through the same analysis pipeline, with checks matched to what each format supports.
A record you can stand behind
Every scan is logged: what was checked, what was found, what could not be established, down to the engine and ruleset version that ran.
Built for teams that defend their decisions.
Insurance claims. Employment screening. Legal disputes. If your document calls have to hold up later, in front of an ombudsman, a regulator, an auditor or a court, Vrfy was built for you.
What Vrfy is not
Vrfy is not a fraud oracle, because no honest tool can be one.
It cannot confirm a document is genuine, and it will never claim to. It cannot tell you whether a real document belongs to the person who sent it. It tells you what it checked, what it found, and what remains unknown.
If a vendor promises you more, ask them to show their working.
Be first in line.
We're onboarding early users in batches. Leave your email and we'll reach out the moment your spot opens up.
Questions
Do you use AI?
The detection engine is deterministic and rules-based. No LLM sits in the detection path, and no third-party AI model receives or processes your documents.
What can it check today?
PDFs first: payslips, bank statements, invoices, receipts and similar business documents, plus scans and photos of them.
When can I use it?
Early access is onboarding in batches. Waitlist members go first, and the first cohort gets founding pricing.