IDV Platform
IDV Platform is a turnkey identity verification service for enterprises — delivering document verification, facial biometrics, liveness detection, and fraud prevention. It is available both as a cloud-hosted SaaS managed by Innovatrics, and as a self-hosted on the customer's own infrastructure for organizations with data residency, compliance, or data ownership requirements.
Key capabilities
| Capability | Description |
|---|---|
| Document verification | Auto-capture and OCR for identity documents from 180+ countries. Extracts data from VIZ, MRZ, QR codes, barcodes, and NFC chips. Cross-checks QR code and barcode data against OCR output. Supports Latin, Cyrillic, Arabic, Thai, Traditional Chinese, and Bengali scripts. Validates document authenticity through logical and visual checks including color profile evaluation. Detects screen-based attacks — a document photo presented on a tablet, smartphone, PC, or TV screen instead of the physical document. |
| Biometric face comparison | Compares the user's selfie against the portrait on the identity document using NIST FRVT top-ranked algorithms. |
| Presentation Attack Detection (PAD) | Multi-layered defence combining passive and active liveness (iBeta Level 1 & 2 certified, ISO 30107-3 Level 2, 100% accuracy in Level 2 testing, 99.9% completion, results under 1 second) and deepfake detection using noise pattern analysis to identify AI-generated and face-swapped images, including from architectures unknown at release. |
| Document NFC chip verification | Reads cryptographically secured data from eMRTD RFID chips (ICAO Doc 9303) in TD1, TD2, and TD3 format — e-passports, national eIDs, electronic driver's licenses, and residence permits. Establishes an encrypted channel via Basic Access Control (BAC, using MRZ data as the key); verifies the issuing government's digital signature via Passive Authentication (PA) to confirm data has not been altered since issuance; proves the chip is not a clone via Active/Chip Authentication (AA/CA) challenge-response. Cross-checks chip data against the presented document and OCR output. Requires no specialized hardware — any NFC-enabled smartphone suffices. |
| Age estimation | Estimates user age and gender from the selfie; validates both against document-stated values. Age estimation ranked #1 in NIST FATE AEV (lowest mean absolute error). |
| Injection attack detection (IAD) | Blocks attempts to bypass the camera by feeding synthetic or pre-recorded media into the capture pipeline — virtual cameras, device emulators, hardware (HDMI) frame replacement, hooked or repackaged apps, and man-in-the-middle interception. The capture components cryptographically sign each captured frame on the device, and the platform verifies that signature server-side, so only genuine, unaltered camera data is accepted. This complements deepfake detection: deepfake analysis alone cannot stop injection, because ordinary (non-deepfake) images can be injected too. |
| Identity duplicity check (1:N) | Searches each applicant against the Customer and Blocklist watchlists to detect duplicates and identity reuse — by face (1:N) and by document data (name, date of birth, personal number). Recognizes returning customers, catches repeat fraudsters, and flags concurrent onboardings. |
| Fraud prevention | Manual review queue for ambiguous cases, configurable Customer and Blocklist watchlists, full audit trails, and session recording. |
| Configurable trust & decisioning | Configurable Trust Factors turn biometric scores, document checks, liveness and age signals into an accept / review / reject decision. Multiple profiles can be defined and assigned per workflow to match each use case's risk appetite. |
| Configurable workflows | Compose verification flows from vetted building blocks — capture steps, checks, decision rules, and outcome actions — without writing code. Fork a predefined workflow or build your own, scoped per tenant. |
| Back-office & case management | Built-in CRM-style back office for operators and supervisors: review queue, Digital Identity and Customer records, duplicate resolution, data correction, and a full audit log, with role-based access. |
| Palm biometrics | Contactless palm enrolment and 1:1 verification as an additional authentication factor — captured via any standard smartphone camera with no specialized hardware; video injection attack detection prevents spoofing. |
How it works
The Identity verification flow is designed to complete in under a minute:
- The user photographs their identity document (both sides where applicable).
- The user takes a selfie.
- Background verification runs automatically — face matching, liveness check, and document authenticity — and returns a result within seconds.
IDV Platform supports workflows for various use cases beyond the standard onboarding flow:
| Workflow | Use case |
|---|---|
| Onboarding | Enrol an unknown person as a verified Customer. |
| Update of personal data | Re-verify an existing Customer to update their document data. |
| Biometric authentication | Authenticate a known Customer by face for transaction approval. |
| Age verification | Verify age of an anonymous user — by facial estimation or document check. |
| Document enrolment | Extract document data without biometric proof (e.g. hotel check-in). |
| Biometric identification | Identify an unauthenticated user by face against the Customer database. |
Target industries
- Financial services — remote banking, transaction authentication
- Telecommunications - client registrations
- Sharing economy — car sharing, marketplaces, freelance and gig platforms
- Travel & hospitality — vacation rentals, contactless check-in
- Healthcare
Integration approach
Standards and certifications
Innovatrics biometric algorithms are independently benchmarked by NIST, with biometric algorithms ranking among top performers since 2004:
| Modality | Benchmark |
|---|---|
| Face recognition 1:N | NIST FRTE 1:N — ongoing evaluation of one-to-many face identification algorithm accuracy and speed across large datasets with varied image quality, pose, and demographics |
| Face recognition 1:1 | NIST FRTE 1:1 — ongoing evaluation of one-to-one face verification algorithm accuracy and speed across large datasets with varied image quality, pose, and demographics |
| Age estimation | NIST FATE AEV — evaluation of automated age estimation from facial images |
| Presentation Attack Detection (PAD) | iBeta Level 1 & 2 — ISO 30107-3 presentation attack detection, 100% spoof detection in Level 2 testing |
| Face image quality | ICAO TR Portrait Quality — face image quality requirements for biometric travel documents |
| GDPR | No personal data retained after session |
License Agreement
Use of this solution is subject to the Innovatrics License Agreement.