How it works

From raw signal to risk decision

DevizeScore turns noisy device and behavior telemetry into a stable fingerprint, risk score, and recommendation your fraud stack can act on immediately.

Request DemoView Docs
  1. 01

    Signals collected

    SDK and server events collect device, network, sensor, and session context.

    App AttestIP addressTouch variance
  2. 02

    Fingerprint resolved

    Persistent identity links the session to known devices and device clusters.

    fp_9f8e7d6cNew device false
  3. 03

    Risk score

    Attestation, behavior, and graph context resolve into one risk score.

    Stability 0.85Risk score 15
  4. 04

    Recommendation

    ALLOWRisk level LOW
    15LOWfinal_risk_score

The trust spectrum

Beyond binary risk decisions

Every device is scored on a continuous trust spectrum. New devices start uncertain. You decide where to draw the line.

0 – 100
LOWMEDIUMHIGHCRITICAL
  • Pixel 7 Pro12 · LOW
  • iPhone 1528 · LOW
  • Galaxy S2452 · MEDIUM
  • Unknown VM78 · HIGH
  • Emulator x8694 · CRITICAL

Sample data

Compliance

  • Hosted in India, multi-region ready
  • DPDP Act 2023 aligned
  • SOC 2 Type II audit in progress
  • GDPRData processing agreements and EU subject rights workflows.In progress
  • ISO 27001Information security management system aligned with ISO/IEC 27001:2022.In progress

Capabilities

Built for depth, designed for simplicity

01 · Persistent Fingerprinting

sess_demo_checkout_42

Hardware-backed identifiers that survive app reinstalls, factory resets, and device cloning attempts.

fp_a1b2c3reinstallfactory resetclonefp_a1b2c3

02 · Risk Signals

sess_demo_checkout_42

Emulator, root, VPN, mock GPS, screen mirroring, MITM.

Emulatoremulator
  • Digital Lending & NBFC78 · HIGH
  • Insurance & InsurTech78 · HIGH
  • Ride-Hailing & Mobility94 · CRITICAL
  • Travel & Hospitality78 · HIGH
  • Social Commerce & Quick Commerce78 · HIGH
  • Bot Defence94 · CRITICAL
  • Identity Defence78 · HIGH
  • Trust & Abuse Defence52 · MEDIUM
  • Ad Fraud Defence94 · CRITICAL
  • Incentive Abuse & Collusion78 · HIGH

03 · Behavioral Biometrics

sess_demo_checkout_42

Accelerometer, gyroscope, touch, keystroke patterns analyzed for session anomalies.

04 · Platform Attestation

sess_demo_checkout_42

Play Integrity, App Attest, DeviceCheck verified server-side.

App Attestvalid

05 · Cross-Platform

sess_demo_checkout_42

One API, every surface — native, hybrid, web, and server-side.

  • Android
  • iOS
  • Web
  • React Native
  • Flutter

06 · Configurable Thresholds

sess_demo_checkout_42

Per-merchant block + step-up, velocity checks, geo anomaly detection.

07 · Real-time decisions

sess_demo_checkout_42
  • 09:3212LOWApp Attest validALLOW
  • 09:4128LOWVPNALLOW
  • 10:0552MEDIUMScreen mirrorREVIEW
  • 10:0778HIGHEmulatorBLOCK
  • 10:1294CRITICALMock GPSBLOCK

Comparison

Where device intelligence changes the operating model

Fraud teams usually choose between brittle rules, point-solution IDV, or a custom signal stack. DevizeScore combines device, behavior, attestation, and identity context behind one risk decision.

CapabilityDevizeScoreRules-based fraudSingle-vendor IDVDIY signals
Persistent device identityBuilt in: Survives reinstall and device reset patternsfingerprint_idis_new_devicestabilityLimited: Usually session or event scopedPartial: Often tied to a verification eventPartial: Possible, but expensive to harden
Behavioral biometricsBuilt in: Touch, motion, and timing signalspointer_eventsaccel_intervalsgyro_intervalskey_press_eventsLimited: Thresholds after behavior is knownPartial: Usually document or selfie focusedPartial: Requires sensor collection and models
Platform attestationBuilt in: App Attest, DeviceCheck, Play Integrityplay_integrity_tokenapp_attest_assertiondevice_check_tokenLimited: Rarely native to rule enginesPartial: May cover app integrity indirectlyPartial: Requires per-platform maintenance
Device farm detectionBuilt in: Clustered devices and emulator patternsis_emulatorcluster_idPartial: Catches known velocity patternsLimited: Not the primary detection surfacePartial: Needs graphing and feedback loops
Configurable enforcementBuilt in: Merchant block and step-up thresholdshard_block_thresholdstep_up_thresholdBuilt in: Strong rule policy controlsPartial: Step-up is usually verification basedPartial: Flexible, but ops-heavy
Cross-platform coverageBuilt in: Mobile, web, and server-side flowsplatformsdk_versionuser_agentPartial: Works where events are instrumentedPartial: Best around onboarding flowsPartial: Requires SDK and backend ownership
Fraud-ring clusteringBuilt in: Links accounts, devices, and behaviorENTITY_REUSE_DETECTEDcluster_idevidence_countPartial: Finds obvious shared attributesLimited: Usually person-check focusedPartial: Needs graph storage and review tooling
Audit-ready decision contextBuilt in: Risk level, scores, signals, recommendationfinal_risk_levelrisk_causesrecommendationPartial: Rule hits explain the policy outcomePartial: Verification artifacts explain identity checksPartial: Only if logging is built deliberately
Built inPartialLimited

Ready to score your first device?

Get started with the DevizeScore API. Explore the documentation or talk to our team about your use case.

Request DemoView Docs