3DiVi Introduces Session Intelligence to Close the Hidden Gap in Biometric Authentication Monitoring

Biometric authentication technologies have reached high levels of accuracy, yet many organisations still struggle with real-world deployment challenges. To address this, 3DiVi, a biometric and Vision AI company, has announced a new analytical approach designed to address what the company calls a critical blind spot in modern biometric identity verification systems—the gap between strong performance metrics and actual operational reliability.

The announcement follows insights shared by Anton Sinkov, Head of the 3DiVi BAF R&D Team, who highlighted why traditional performance indicators often fail to reflect how biometric authentication behaves in production environments.

The Problem: When “Good” Metrics Hide Real Risks

Facial biometric authentication has been widely adopted across digital onboarding, KYC, and identity verification workflows for more than a decade. Accuracy of face matching and anti-spoofing technologies has significantly improved, allowing systems to reliably detect attempts involving photos, videos, or masks.

However, according to 3DiVi, operational maturity has not kept pace with algorithmic progress.

Most biometric deployments rely on standard indicators such as:

• APCER: Cases where spoof attacks are incorrectly accepted.
• BPCER: Genuine users mistakenly rejected.
• Conversion rate: Successful verification completion.

When these values remain within expected ranges, systems are typically considered healthy.

Yet in practice, organisations frequently experience user losses, hidden instability, or localised failures despite “normal” metrics.

“The industry evaluates biometric authentication systems based on what they already measure,” said Sinkov. “Metrics provide a snapshot, but they do not show how system behaviour evolves over time or where risks are accumulating.”

3DiVi Bridges a Structural Industry Gap

3DiVi emphasises that the issue is not tied to individual implementations but reflects a broader market structure.

Typically:

• Vendors provide algorithms and baseline metrics.
• Integrators configure authentication workflows.
• Customers monitor dashboards.

What remains missing is a dedicated monitoring layer capable of analysing biometric decisions after pass/fail outcomes occur.

As a result, critical operational problems often remain invisible, including:

• Rising false rejections caused by declining image quality.
• Environmental changes affecting specific devices or locations.
• Gradual behavioural drift within deployed models.
• Attack patterns that remain hidden behind aggregated statistics.

In one example cited by the company, lighting changes introduced by a background advertising screen at a single KYC terminal reduced image quality. While users experienced increased rejection rates, aggregated system metrics showed only minor deviations and failed to flag the issue.

The Solution: Session Intelligence

To address this gap, 3DiVi introduced Session Intelligence, a dedicated analytical layer integrated into  3DiVi BAF, a biometric identity verification platform for digital identity systems in banks, fintechs, and government services.

Rather than evaluating only aggregated statistics, Session Intelligence analyses real authentication sessions over time.

The approach includes:

• Post-analysis of real registration and authentication attempts.
• Identification of rejection causes and module behaviour.
• Detection of algorithmic inconsistencies.
• Monitoring of attack frequency and evolution.
• Evaluation of environmental and device-level impacts.

According to 3DiVi, many operational risks exist exclusively at the session level and never appear in traditional telemetry.

What Session-Level Analysis Reveals

Session Intelligence enables organisations to detect patterns typically missed by dashboards, including:

• Systematic rejection of valid users due to specific quality checks.
• Instability where images without visible issues still fail verification.
• Emerging attack strategies that remain statistically invisible.
• Configuration or environmental factors influencing performance.

By formalising session analysis as an ongoing process, the system transforms biometric monitoring from static evaluation into continuous behavioural observation.

Enterprise Outcomes Without Workflow Disruption

For enterprise customers, the monitoring layer produces periodic analytical reports documenting:

• Current performance indicators.
• Identified operational risks.
• Attack examples and trends.
• Root-cause analysis of errors.
• Optimisation recommendations.

Importantly, the analysis runs asynchronously on accumulated data and does not affect onboarding speed, require manual verification, or introduce additional operational overhead.

From Black Box Metrics to Operational Visibility

3DiVi positions Session Intelligence as a shift in how organisations interact with biometric authentication systems.

Without advanced monitoring, biometric platforms often function as black boxes validated only by acceptable metrics. With session-level analysis, enterprises gain visibility into real system behaviour over time.

“The key change is moving from trusting indicators to understanding behaviour,” Sinkov explained. “Organisations no longer evaluate isolated numbers—they observe how authentication actually performs in live operation.”

With Session Intelligence, 3DiVi aims to redefine biometric authentication management by adding continuous operational awareness to already mature recognition technologies—helping enterprises detect risks earlier, maintain conversion rates, and ensure long-term system stability.