Security-first architecture for irreversible biometric identity.
PalmPass is engineered to protect user privacy through data minimization, irreversible vector representations, and strict cryptographic token isolation.
Current Prototype vs Production Security Roadmap
We strictly separate current MVP controls from our production security architecture. Current software implementations utilize cryptographic vector hashing and simulated session boundaries; formal third-party HSM validation, SOC 2 certification, and ISO biometric audits are part of our upcoming validation roadmap.
Comprehensive security across every layer of the architecture.
Data minimization, irreversible vectors, and lifecycle security controls.
Privacy & Data Minimization
Privacy is not a feature layered on top; it is the fundamental architectural constraint of PalmPass.
Data Minimization
Implemented in MVPRaw palm photographs are processed in volatile memory and permanently discarded immediately after vector generation.
Explicit User Consent
Implemented in MVPEnrollment requires active user initiation with transparent disclaimers on intended organizational use.
Purpose Limitation
Production ArchitectureAuthentication scopes restrict token validity exclusively to the authorized requesting application.
Biometric Protection
Biometric vectors are mathematically irreversible, meaning the original palm image cannot be reconstructed.
Protected Representations
Implemented in MVPBiometric templates are stored as irreversible cryptographic vector embeddings, never as raw images.
Tokenized Identity Verification
Implemented in MVPExternal systems receive short-lived identity tokens rather than biometric data or persistent identifiers.
Cryptographic Revocation
Production ArchitectureIf an identity record is compromised or an employee departs, templates can be permanently revoked and invalidated.
Authentication Security
Multi-layered defense against spoofing, presentation attacks, and unauthorized device capture.
Active Liveness Detection
Implemented in MVP (Iterating)Analyzes surface reflectance and micro-texture variations to detect photos, video replays, and silicone molds.
Device Integrity Verification
Production ArchitectureCapture devices must authenticate using hardware keys before submitting scan sessions to the verification engine.
Access Control & Rate Limiting
Production ArchitectureStrict per-device and per-user velocity limits prevent automated brute-force authentication attempts.
Infrastructure & Network Security
End-to-end cryptographic protection across the entire transmission and verification lifecycle.
Encryption in Transit & at Rest
Implemented in MVPTLS 1.3 encryption with strict cipher suites in transit; AES-256-GCM encryption for stored vector databases.
API Authentication & Signing
Production ArchitectureMutual TLS (mTLS) and cryptographically signed request payloads between client endpoints and PalmPass.
Audit Logging & Tamper Evidence
Production ArchitectureEvery verification request records an immutable cryptographic event entry for enterprise compliance.
Continuous Monitoring & Anomaly Detection
Production ArchitectureAutomated telemetry flags anomalous retry frequencies or distributed presentation attack patterns.
Identity Lifecycle Security
Rigorous protocols governing every stage of a user's biometric relationship with PalmPass.
Secure Enrollment
Implemented in MVPMulti-factor supervised verification when linking an individual's palm to their enterprise account.
Sub-Second Authentication
Implemented in MVPSingle-presentation verification with ephemeral token lifecycle limited to 60 seconds.
One-Click Revocation
Production ArchitectureInstant template deletion and cryptographic salt disposal upon user offboarding.
Seamless Re-Enrollment
Production ArchitectureSecure, audit-logged replacement of biometric representations when updating hardware models.
Frequently asked security & privacy questions.
Clear, technically accurate answers on our biometric handling.
Can an attacker reconstruct a picture of my hand from the PalmPass database?
No. PalmPass converts optical scans into irreversible numerical vectors (mathematical embeddings). The transformation is one-way, meaning it is mathematically impossible to reconstruct the original palm image or fingerprint ridges from stored template data.
How does PalmPass prevent fake hands, photographs, or silicone molds?
PalmPass incorporates multi-parameter liveness detection. The system analyzes skin texture reflection, high-frequency noise variance, and spatial geometry to ensure a real, living human palm is present before processing biometric matching.
Is PalmPass compliant with GDPR and biometric privacy laws?
PalmPass is architected specifically around the core principles of GDPR Article 9 (special category biometric data): data minimization, explicit consent, purpose limitation, and the right to erasure (cryptographic revocation).
Have specific security or compliance requirements?
Connect with our engineering team or explore an early access pilot evaluation for your organization.