A branded card in weeks, with onboarding your compliance team runs
Virtual cards tokenized to Apple Pay and Google Pay on day one, KYC flows configured per programme, Fraud AI on every tap, and disposable cards for subscriptions and privacy.
FRAUD AI · IN-LINE, EVERY AUTH
Block the 0.97s.
Wave through the 0.03s.
Adaptive models score every authorization in 18ms, inside the authorization flow, not after it. Card testing, cloning, first-party abuse and account takeover are caught before the response leaves Rigid.
- Score latency
- 18ms
- Score latency
- Fraud loss rate
- 0.02%
- Fraud loss rate
- Of authorizations scored
- 100%
- Of authorizations scored
rigid · risk engineScoring in-line
auth · **** 1180 · $1.00 · e-comrisk score
0.000.501.00
velocity · 14 auths / 60s+0.41
amount pattern · $1.00 probe+0.32
network · merchant reputation+0.18
history · known holder pattern−0.02
Decline · card_testingdecided in 18ms
HOW IT DECIDESTrained on the whole network.
Trained on the whole network.
Tuned to your programme.
SIGNALS
400+ features per auth
Velocity, geography, merchant history, amount patterns and network reputation, computed in-line for every authorization.
ADAPTIVE
Learns per programme
A travel card and a teen card have a different normal. Models adapt to each programme's traffic without hand-written rules.
GRADUATED
Decline is the last resort
Approve, step up to 3-D Secure, refer for review, or decline. Graduated responses keep good customers moving.
EXPLAINABLE
Every score, itemized
Signal-level contributions on every decision, recorded in the decision trace for your risk team, your auditors and your cardholder support.
YOUR RULES TOO
Rules where you want them
Hard limits, geo-blocks and merchant-category policies run alongside the models and are tuned per programme without a deploy. Deterministic where regulation demands it.
FEEDBACK LOOP
Verdicts close the loop
Reviewer decisions and confirmed fraud feed straight back into training. The model improves with every case your team closes.
WHO BUILDS ON IT
Who Fraud AI is for, and what they get.
The same platform, used differently by different teams. Each card names the segment, the outcome, and the products that carry it.
Many client programmes, one console, one API
Each programme is a native object with its own rules, ledger and flows. Launch a new card product or a new market as configuration, and see every programme's books in one trial balance.
Keep your ledger and your decision, replace the network side
Gateway mode asks your endpoint for every authorization and posts instructions back to your books. Stand-in processing keeps cards working inside limits you declare when your core is unreachable.
One virtual card per booking, in the merchant's currency
Time-boxed, MCC-limited cards for every hotel and flight, FX applied once at authorization across 150+ currencies, and fraud scoring tuned to travel's spiky, cross-border traffic.
Run your last month's traffic through it.
We replay your historical authorizations against Fraud AI and show you what it would have caught.