Use cases · Marketplace & e-commerce
Trust the two sides. Decide every transaction.
Marketplaces face risk from both sides of the listing — buyer abuse, seller fraud, payment fraud, promo abuse, and reputation manipulation. Centurion plugs into every point of that journey, on a single decisioning fabric, with every decision audit-logged.
By customer journey
Five lifecycle stages. Every dollar of marketplace risk flows through them in order.
From buyer acquisition to seller payout — Centurion plugs into every stage where marketplace risk gets decided. Find your stage. Find your use case.
01
Attract
Acquire buyers & sellers
- Seller-acquisition lead scoring
- Buyer-acquisition channel risk
- Affiliate / referral channel quality
- Marketing-spend risk weighting
02
Acquire
Sign up & onboard
- Buyer account opening
- Seller onboarding & KYC / KYB
- Document & business verification
- Mule-seller and shell-storefront detection
- Synthetic-identity defense at signup
03
Engage
Authenticate, list & transact
- Login & session-trust scoring (ATO, bots)
- Listing creation risk (counterfeit, off-platform, illicit)
- Buyer-seller messaging-fraud signal
- Checkout & payment authorization
- Cross-border transaction screening
04
Grow
Service & expand
- Promo, coupon & referral-abuse defense
- Loyalty-point cash-out detection
- Seller-velocity & wash-trading detection
- Fake-review-ring and reputation manipulation
- Seller-account takeover (storefront hijack)
- Account-change risk (payout details, banking)
05
Resolve
Disputes & recovery
- Friendly-fraud & policy-abuse classification
- Refund-abuse and returnless-refund exploit
- Chargeback representment
- Triangulation-fraud unwinding
- Seller payout hold / claw-back routing
The fraud surfaces
Eight surfaces every marketplace has to defend.
Marketplaces concentrate risk in ways single-sided platforms don't. The defense has to read the buyer, the seller, the listing, and the payment — together, at decision time.
Listing fraud
Counterfeit goods, off-platform redirection, illicit categories, brand-jacking. Surface listings before they convert — graph signal across seller, image, IP, and language pattern.
Seller account takeover
Storefront hijack via credential stuffing, session replay, payout-detail change. Behavior + device + telephony coherence catches the changeover before funds redirect.
Fake-review rings
Coordinated review manipulation across buyer cohorts. Population-level graph detection — identifier overlap, behavioral homogeneity, timing clustering across reviewers.
Promo & coupon abuse
Sign-up bonus farming, coupon stacking, referral self-loop, loyalty-point cash-out. Cohort-level identifier-overlap scoring at redemption time.
Refund & return fraud
Refund abuse, returnless-refund exploit, return-of-empty-box, friendly-fraud chargebacks. Disputed-transaction classification with behavioral-trajectory signal.
Triangulation & payment fraud
Card-not-present fraud, triangulation rings using stolen cards through marketplace listings, BIN-attack pattern, gift-card laundering.
Buyer-side abuse
INR (item-not-received) abuse, SNAD (significantly-not-as-described) gaming, dispute-velocity rings. Buyer-trust scoring with cross-merchant consortium signal.
On-platform value exchange
Money laundering via marketplace listings, gift-card recycling, off-platform value transfer disguised as listings. AML-grade transaction monitoring.
What plugs in
All four Centurion components — wired together, tuned to the marketplace shape.
ML Graph
Buyer-seller-payment-device graph at real-time decision latency. Ring detection across listings, reviews, and payment instruments.
Learn more →HD Trust
Session-trust at login, listing creation, and checkout. Bots, stealth browsers, ATO — caught at the moment, not in case review.
Learn more →Model Foundry
Buyer-trust models, seller-risk scoring, dispute-classification recipes. Tuned to your marketplace category mix and GMV profile.
Learn more →FlowX
The decisioning fabric — listing-approval, payment-authorization, payout-hold, dispute-routing strategies in one versioned, A/B-testable layer.
Learn more →What changes
Higher buyer trust. Lower seller fraud. Fewer chargebacks. The same approval rate on legitimate transactions — without the manual-review queue you've been building around.
For marketplace risk leaders
See it on your traffic.
We'll run a benchmark on a slice of your historical listings, transactions, and disputes — quantified lift, before you ship anything.