Fraud detection
Catch synthetic identities, mule accounts and coordinated rings before money moves. Behavioural, device and network signals resolved in one call.
Velocity rulesLink analysisMule detection
Identify. Score. Secure.
Powering fraud prevention, credit intelligence and digital trust for modern businesses.
Trusted by risk and growth teams across regulated industries
Products
Eighteen risk products.
One integration.
Start with the check you need today. Add the rest without touching your integration — every product resolves into the same decision object.
Credit decision engine
Write policy once, run it on every application. Approve, refer or decline with a full reason trail.
Policy versioningReason codes
Document verification & OCR
Read, validate and tamper-check documents in 40+ formats with field-level confidence.
Device & mobile risk intelligence
Fingerprint every session. Detect emulators, rooted devices, SIM swaps and recycled numbers before onboarding completes.
Emulator checkSIM swapNumber ageGeo drift
Face match & liveness
Match a selfie to an ID photo with passive liveness. Deepfake and replay attacks flagged in the same response.
Estimated income
Infer income bands from bank, employment and alternate data — without asking for another payslip.
Ad fraud detection
Separate real installs from bots, click farms and attribution fraud before you pay for them.
Bot trafficClick farmsInstall fraud
KYC, KYB & AML screening
Onboard individuals and businesses, then screen against global sanctions, PEP and adverse media lists on a schedule you set.
AadhaarPANBank accountGST & CINSanctionsPEPAdverse media
How it works
Four moves,
one round trip.
STEP 01
Identify
Resolve the person or business behind the application from documents, biometrics and government sources.
STEP 02
Analyze
Enrich with device, email, mobile, bank and behavioural signals, then weigh each one against your policy.
STEP 03
Score
Models resolve the signals into a single risk score, with the reasons that drove it ranked by contribution.
STEP 04
Secure
Return approve, review or decline — logged, versioned and exportable for your next audit.
Government IDpassport · MRZ parsedReadingVerified
Face match & liveness98.6% · passiveMatchingVerified
PAN & Aadhaarsource-verifiedCheckingVerified
Bank accountpenny-drop · name matchPendingVerified
Device integrityclean
Email domain age6y 3m
Mobile number age4y 2m
Velocity across network2 / 30d
Estimated income band₹18.4L
AML & sanctionsno hits
0Risk score
Income band confirmed+0.24
Device history clean+0.18
Mobile tenure high+0.11
Approved
Decision returned in 112 ms with every input, model version and reason code written to the audit log.
Developers
Built by engineers
who ship risk.
One REST endpoint, typed SDKs and a sandbox that returns realistic fraud patterns. Key to first production call in an afternoon.
- SDKs for Node, Python, Java, Go, PHP and Kotlin
- Sandbox keys seeded with fraud and edge-case identities
- Idempotency, webhooks and versioned responses on every route
- Status page, per-route latency and a real human on support
// one call, every signalconst decision = await iscoreit.risk.score({ reference_id: "app_9F2KQ4", identity: { pan: "ABCDE1234F", aadhaar: "xxxx-xxxx-4417", mobile: "+91 98••• ••821" }, document: { type: "passport", file: buffer }, selfie: { file: selfie, liveness: true }, device: { fingerprint: req.device_id }, policy: "lending.retail.v4"});{ "outcome": "approve", "risk_score": 94, "identity": { "pan": "verified", "face_match": 98.6, "liveness": "passed" }, "signals": [ { "code": "device.clean", "w": 0.18 }, { "code": "income.confirmed", "w": 0.24 } ], "latency_ms": 112}0Median latency
0Live endpoints
0Uptime
AI intelligence
Models that learn
from every decision.
Signals in
Every call arrives as raw inputs — documents, biometrics, device, email, mobile, bank and behavioural history.
Feature resolution
Hundreds of derived features are computed per application, normalised across geographies and product types.
Ensemble inference
Gradient-boosted models vote alongside graph and sequence models, retrained weekly on realised outcomes.
Decision out
One score, one outcome, and the ranked reasons behind it — the same contract on every route.
0API calls processed
0Platform uptime
0Response latency
0Verification APIs
0Enterprise customers
Industries
Different risk.
Same infrastructure.
Security & compliance
Audit-ready from
the first call.
Encrypted in transit and at rest, isolated per tenant, retained only as long as your policy says. Every decision is logged with its inputs, model version and reason codes.
ISO 27001Information security
SOC 2 Type IIAudited annually
DPDP readyConsent & erasure flows
RBI alignedIn-region data residency
AES-256Per-tenant keys at rest
PCI DSSCard data handling
Customers
What risk teams say.
Ready to build trust
at scale?
Tell us what you're onboarding and we'll run the decision flow on your own data in a 30-minute working session.