Tamperlens and Inscribe

How to read this page

Side by side

One design decision, seen from both ends

Where Inscribe is the better answer

• You are a bank, credit union, lender or fintech with a fraud team. That is who the product is built for, in their own words, and the analyst-queue triage assumes you have analysts to route to.
• Network signals. A template reused across their customer base is a signal that only exists because they retain and pool what they see. We keep nothing, so we can never have it, by design, and it is a real trade.
• Your stack is Plaid, Alloy, Taktile. They integrate with it (read 2026-08-13); we do not.
• Procurement asks for certifications. SOC 2 and ISO 27001 against our none.
• The company they keep. Plaid, Ramp, Bluevine, Mercari and Navan on the logo wall (read 2026-08-13) is a different scale of vetting than a solo-operated product can point to.

Where Tamperlens is the better answer

• You are not in financial services. Their positioning names their buyer. An imobiliária, an HR team, a law office, a marketplace (anyone outside that vertical, or without a procurement process) has no published path into Inscribe. That buyer is who this page is for.
• You want a price before a call. Ours is at /pricing, in USD and in reais. Theirs cannot be established from outside (read 2026-08-13).
• You want to check one document today. The free checker takes a file with no signup; an API key comes with 50 free documents a month.
• You need the check inside CI. The same bytes produce the same report, so results can be regression-tested. A property a learning system does not promise.
• You want the error rate before you buy. Our false-positive rate is at /evidence with its population and date, and the engine's release log is public at /changelog.
• Nothing may be retained. Documents are uploaded directly, parsed in memory and never written to disk: the flip side of not having network signals.
• Your files are not bank statements. Any PDF, image or Office file goes through the same endpoint, including redaction recovery and prompt-injection signals.
• Brazil. CPF, CNPJ and boleto identifiers checked by arithmetic on the check digits, with prices quoted in BRL.

What we cannot do

• No network signals. We keep nothing between requests, so a template reused across companies is invisible to us.
• No analyst queue. We emit signals and let your policy decide; we ship no review UI and no human review service.
• No bank-data connections and no identity verification. We never look past the file.
• No SOC 2 or ISO 27001, disclosed at /security, and no on-premise deployment.
• Recall is unmeasured. We know how often we flag genuine documents; we do not know how often we catch real tampering.
• One region. Hosting is EU-only, and this is a solo-operated product.

The number a buyer actually needs

Inscribe declining to publish an unmethodologised accuracy number is the right instinct, and rarer than it should be. It is still not the same as publishing the number that tells a buyer what review will cost: how often the system flags a genuine document. Neither Inscribe nor almost anyone in this market publishes that (read 2026-08-13).

Ours is at /evidence, with the population it was measured on, the date, and the engine version that produced it. The other comparisons are at /alternatives.