Trust Stamp Detect

High-performance deduplication

To protect your systems from bad actors using multiple fraudulent documents or falsified identifiers, it is critical to ensure that each person only has one unique profile. Biometrics present a transformative way to determine that every new applicant is unique from all existing users. A bad actor cannot change their name or identifiers to create multiple accounts, because their face is unique to them.

However, using biometric data captured on enrolment to deduplicate presents multiple challenges for an organisation with many users and applicants:

1. Storing large amounts of biometric information puts sensitive data at risk
2. Computational inefficiencies in matching biometric templates of all applicants with all existing users lead to extensive cost and time burdens

However, using biometric data captured on enrolment to deduplicate presents multiple challenges for an organisation with many users and applicants:

The next level of fraud prevention

Trust Stamp's proprietary data transformation and deduplication technology dramatically reduces the processing power needed to deduplicate large volumes of records accurately, saving time and money while augmenting your fraud-prevention strategy with a robust matching protocol that has not been computationally feasible until now. Seamlessly analyse your entire user base on every new entry for the next level of fraud prevention.

Benefits

Efficient, tailored solutions that last

Prevent Fraud and Protect Data

Biometric deduplication is a powerful indicator of synthetic identities. Trust Stamp's transformation algorithms minimise and obfuscate raw information, turning PII into pseudonymised data that serves no use if compromised. With Trust Stamp Detect, you can determine the uniqueness of one identity among many, without sharing or revealing sensitive information

Learn More

Leverage Unmatched Efficiency

Trust Stamp’s patented transformation and comparison technologies significantly minimise the amount of data necessary for accurate deduplication, and probabilistic matching protocols limit analysis to only the most likely matches. These factors combined make large-scale biometric deduplication computationally possible, operationally realistic, and financially attractive.

Technical Whitepaper

Adjust to Meet Risk-Acceptance Needs

Adjustable matching thresholds allow your organisation to set criteria based on risk-acceptance needs, and to change thresholds on-the-fly to balance good friction with security requirement

Build Robust Systems with Longevity

Fraud-prevention functionality increases as your user base grows, without ever creating a honeypot of biometric data for bad actors to target. The addition of a randomising seed makes it possible to revoke tokens in the event of compromise

How it works

Advanced AI and machine learning techniques

Detect attempted attacks

Multiple anti-spoofing tests detect print and device display attacks

Validate document attributes

AI matches proprietary templates and analyses security features for authenticity

Check for tampering

Comparison of visual text and machine readable text determines document tampering

Go further with facial biometrics

When combined with facial biometrics, the document photo is matched against a live facial capture to establish your user's true identity

Learn More

The core technology

Trust Stamp’s identity proofing and deduplication technology maximises the privacy and utility of identity information by leveraging pseudonymised token identifiers for accurate, low-resource matching within your systems. If ever compromised, tokens provide no value to would-be identity thieves