One fraud report protects the whole network.
Adoor is Masenu's first product — shared fraud intelligence that lets telcos, fintechs and banks trade confirmed fraud signals on hashed identifiers and get a real-time risk decision before money moves. A fraudster gets one shot per network, not one per institution.
Identifiers hashed at your edge — raw data never reaches Adoor
Built for institutions, individuals, and anyone about to send money.
Contribute confirmed fraud reports, get sub-150 ms risk decisions on lookups, and see the cross-network rings no single institution can spot alone.
Explore membershipProtect your number, wallet or ID. When a business checks a protected identifier, they verify more carefully — and you see who checked.
Protect yourselfAbout to send money to a number you don't know? Check it against consortium fraud intelligence first. No account needed.
Check a numberOne identifier in. A network-wide decision out.
No raw phone number, account or ID ever leaves a member's systems. Only one-way keyed digests collide across the network.
Members normalize and HMAC-hash identifiers inside their own systems. Only keyed digests reach Adoor — never a raw phone number, account or ID.
The same real-world identifier produces an identical digest for every member, so confirmed fraud collides mathematically — and a graph engine links related digests into rings.
Before onboarding or payout, one API call returns a scored, explainable decision — allow, step-up, review or block — engineered for p95 under 150 ms.
One identifier. Two peppers. Zero exposure.
The same real-world identifier deterministically produces the same key across the whole network — so confirmed fraud collides — while nothing stored can be turned back into a phone number, account or ID. It's the sequence the panel above steps through.
A raw phone, wallet, account or ID — inside your systems. It never leaves; Adoor never receives it.
Your SDK normalizes and HMAC-SHA256s the value with the shared consortium pepper. The same identifier makes the same h1 for every member — that's what lets hashes join.
Adoor applies a second, platform-only pepper server-side. h2 is the stored join key — it never leaves the server, and can't be reversed to raw PII.
Same input, same hash — so the network joins reports without anyone sharing raw data.
h2 is a keyed digest; no stored value can be turned back into a phone number or account.
h1 needs your consortium pepper; h2 needs Adoor's central pepper. Neither side ever holds both.
A ring is invisible to one institution — and obvious to the network.
Each member sees only its own fragment of the fraud. Correlated across the consortium, the pattern that no one could see alone becomes a detected ring.
One coordinated fraud ring — seven identifiers across three sectors — that no member reported in full.
Neutral by architecture, not by promise.
Adoor never holds raw PII, so it cannot leak, sell, or be compelled to produce it. Contributor identity is reduced to sector + country — reporting never advertises a breach.
Every risk score carries human-readable reasons and a versioned ruleset. Anyone flagged can dispute; contested reports stop driving decisions immediately.
Every lookup, report and admin action lands in an append-only, hash-chained audit log. Built SOC 2-ready from the first commit.
Had your identity used by fraudsters before?
Guardian puts a protective marker on your identifiers — extra verification for anyone using them, and a log of who checked. Fraud victims get Shield free.