Feature comparison

Key differences between vulnerability detection approaches.

Feature EchoDepth Sentiment Analysis Traditional QA Self-Declaration
Coverage 100% 100% 2-5% Self-selected
Detection method Physiological signals (FACS, VAD, prosody) Keyword/phrase matching Manual review Customer disclosure
Timing Real-time Real-time Post-interaction At disclosure
Detects masked distress Yes No Sometimes No
Modalities Video, voice, text, image Text only Varies Verbal/written
FCA audit trail Automatic Partial Manual Manual
Pre-decision intervention Yes Sometimes No No
Hardware required Standard camera/mic None None None

EchoDepth vs Sentiment Analysis

Sentiment analysis reads words. EchoDepth reads people.

Sentiment misses masked distress

A customer saying "I'm fine, let's proceed" while displaying facial stress markers will be classified as neutral or positive by sentiment analysis.

EchoDepth detects physiological signals

44 facial Action Units, voice prosody, and VAD scoring reveal emotional state regardless of words spoken. Involuntary signals are harder to mask.

EchoDepth vs Traditional QA

QA finds problems after harm occurs. EchoDepth prevents them.

QA samples 2-5% of interactions

Most vulnerable customer interactions are never reviewed. Issues are discovered through complaints, not proactive detection.

EchoDepth analyses 100% in real time

Every interaction is scored. Vulnerability flags appear before decisions are made, enabling intervention rather than remediation.

EchoDepth vs Self-Declaration

Vulnerable customers often don't self-identify.

Declaration requires awareness and willingness

Many vulnerable customers don't recognise their vulnerability, feel shame about disclosure, or fear it will affect their application outcome.

EchoDepth detects without disclosure

Physiological markers of distress and cognitive load are detected automatically, regardless of whether the customer chooses to disclose.

EchoDepth vs Generic Emotion AI

Built for regulated financial services, not general purpose.

Generic tools lack FCA compliance features

Consumer emotion AI is designed for marketing, not regulatory compliance. Audit trails, vulnerability tiering, and Consumer Duty evidence are absent.

EchoDepth is FCA Regulatory Sandbox tested

Purpose-built for UK financial services. Consumer Duty audit trails, vulnerability tiering, GDPR-compliant architecture. ISO 9001 infrastructure.

Why EchoDepth

Key differentiators

FACS-based, not keyword

44 facial Action Units tracked per frame based on the Facial Action Coding System. Involuntary physiological signals that can't be masked or scripted around.

Voice-first for phone

Collections and complaints happen on the phone. EchoDepth's prosody analysis works with audio-only interactions.

FCA Regulatory Sandbox

Tested within the FCA's regulatory innovation programme. Built for UK financial services compliance from the ground up.

No specialist hardware

Works with existing webcams, phone cameras, and audio systems. No wearables, sensors, or capital expenditure.

Pre-decision flags

Vulnerability signals appear before loans are declined, accounts closed, or recovery actions initiated. Intervention, not remediation.

Audit trail automatic

Consumer Duty evidence generated automatically. Timestamped records that vulnerability was assessed before every decision.

See the difference

Book a discovery call to see EchoDepth analyse a sample interaction — and compare the output to your current approach.

Book Discovery Call How It Works
Article 22 UK GDPR — Human Review Required: EchoDepth vulnerability and distress signals are advisory outputs for qualified human review — not automated decisions. All vulnerability classifications, product suitability assessments and escalation decisions must be made by a qualified professional. Privacy policy · Terms