Real-time AI scam-call shield

Detect voice scams before they start

Verity fuses deepfake detection, social-engineering analysis, and speaker verification into a single trust score -- in under two seconds.

<2s

Detection latency

5

ML models

0

Models trained

8

Languages

The Problem

AI voice clones are undetectable to the human ear

A 30-second voice sample from a WhatsApp note or social media video is enough to clone anyone's voice. Scammers use these clones to impersonate family members, demanding money while the victim hears their loved one's voice begging for help.

$1B+stolen from families via voice scams in 2025
30sof audio needed to clone any voice
77%of victims could not tell the clone from the real voice
400%increase in AI voice scam reports year-over-year

Typical scam call transcript

"Mom, please, I've been arrested and I need ten thousand rupees right now. Don't tell dad. The police won't let me go until I pay this. Please hurry."

AI CloneGenerated from a 30-second voice sample

Architecture

Three signals, one verdict

Each layer runs independently. The Trust Engine fuses all outputs into a single 0-100 score. If any signal fires, the score drops. Defense in depth -- one layer failing doesn't matter.

01

Deepfake Detection

Wav2Vec2 + AASIST3 neural networks analyze raw audio to detect synthetic speech, voice cloning, and TTS artifacts. Gemini 2.5 provides noise-robust analysis.

02

Scam Pattern Analysis

Gemini 2.5 Flash examines the live transcript for social engineering -- urgency, financial requests, authority threats, impersonation, and secrecy pressure.

03

Speaker Verification

ECAPA-TDNN compares the caller's voiceprint against your Family Voice Vault. If they claim to be your son, the voice must match.

Live Output

One score. Full transparency.

The Trust Score starts at 100. Each red flag subtracts points: synthetic voice detected, financial request identified, voiceprint mismatch confirmed. You see exactly why a call is flagged.

70-100Trusted
40-69Suspicious
0-39Scam detected

100

Safe

Synthetic voice94%
Scam patterns91%
Voice match23%

Real Scenarios

See Verity in action

Four real-world scenarios showing how the three detection layers work together. Try them yourself in the dashboard's Demo Scenarios panel.

Grandparent Scam

6

A scammer clones your grandson's voice from a WhatsApp note and calls demanding bail money.

Detection

Deepfake detector flags synthetic voice at 90%+. Scam classifier catches urgency and financial demand. Voice vault shows no match to enrolled grandson.

IRS Impersonation

12

Automated call threatens arrest for unpaid taxes, demands payment via gift cards.

Detection

LLM detects authority threats (95%) and financial requests (90%). Trigger phrases flagged: 'warrant', 'arrest', 'gift cards'.

Bank Fraud

15

Fake bank rep claims your account is compromised, asks you to transfer money to a 'safe account'.

Detection

Financial request (92%) and impersonation (80%) flagged. Secrecy pressure detected: 'don't share this information'.

Legitimate Family Call

98

Your actual son calls to chat about weekend plans.

Detection

All scam signals at 0%. Deepfake probability 1%. Voice vault match: 95%. Trust Score: 98.

Platform

Everything you need

Not just detection -- a complete protection platform with forensic reporting, education, threat intelligence, and family safety tools.

Live Mic Analysis

Stream audio from your microphone for real-time call monitoring

File Upload

Upload WAV, MP3, M4A, OGG files for instant analysis

Family Voice Vault

Enroll family voiceprints. Verify callers against trusted contacts

Scam Alert Overlay

Full-screen emergency warning with action buttons when scam detected

Forensic Reports

One-click export of HTML reports and raw JSON analysis data

AI-Powered Reasoning

Gemini explains why a voice sounds synthetic or a call seems suspicious

Demo Scenarios

4 pre-built scenarios to demo the system without a backend

8 Languages

English, Hindi, Spanish, Tamil, French, German, Japanese, Chinese

Privacy-First

No audio leaves your device. Voice Vault stays local. Zero cloud storage

Threat Intelligence

Community feed showing active scam campaigns and trending attack types

Scam Education

Interactive library of scam playbooks with a quiz to test your knowledge

Emergency Contacts

Quick-call buttons for trusted family members when a scam is detected

Voice Vault

Aarav

Son

98%

Priya

Daughter

96%

Mom

Mother

94%

Protection

Family Voice Vault

Enroll a 30-second voice sample of each family member. When someone calls claiming to be your son, Verity checks the voiceprint against the vault -- exposing clones that sound identical to the human ear.

The vault uses ECAPA-TDNN to create 192-dimensional voiceprint embeddings stored locally on your device. No biometric data ever leaves your phone.

Open Voice Vault

Landscape

Why not the existing players?

Scam-call protection exists -- for telecom networks, banks, and browsers. Nobody protects the consumer from AI voice clones. Until now.

SolutionProtectsGap
HiyaTelecom networksB2B only, no deepfake detection
PindropBanks & call centersEnterprise pricing, no consumer app
McAfeeBrowser scam alertsNo real-time voice analysis
VerityYour family's phoneAll three signals, free, on-device

Under the Hood

5 models. Zero training. Ship-ready.

Every ML component is a pretrained downloadable artifact. We built the fusion and the product, not the models.

Audio Forensics

Gemini 2.5

Noise-robust deepfake + scam + transcript in 1 call

Deepfake Detection

Wav2Vec2

XLS-R 300M fine-tuned on ElevenLabs

Anti-Spoofing

AASIST3

KAN + GAT for voice conversion attacks

Speaker Verify

ECAPA-TDNN

192-dim voiceprint embeddings

Transcription

Whisper

distil-large-v3, 8 languages

FAQ

Frequently asked questions

How does Verity detect AI-cloned voices?+

We use three layers: a Wav2Vec2 model trained specifically on ElevenLabs and modern TTS, AASIST3 for older voice conversion attacks, and Gemini 2.5 Flash for noise-robust audio forensics. If any layer detects synthesis, the trust score drops.

Does it work with real human scammers (not AI)?+

Yes. Even when the voice is real, our LLM analyzes the transcript for social engineering patterns -- urgency, financial demands, authority threats, secrecy pressure. Real human scam calls are caught through content analysis.

What is the Family Voice Vault?+

You enroll 30-second voice samples of trusted contacts (family members). When someone calls claiming to be your son, Verity's ECAPA-TDNN model compares the caller's voiceprint against the enrolled sample. A clone may sound perfect to your ear, but the voiceprint won't match.

Is my audio data stored or shared?+

No. Audio analysis happens on-device or via encrypted API calls. No audio is stored on our servers. The Voice Vault is local to your device.

What languages are supported?+

Transcription supports English, Hindi, Spanish, Tamil, French, German, Japanese, and Chinese. The scam classifier works across all languages since it analyzes translated content.

How fast is the analysis?+

Under 2 seconds for real-time streaming. Uploaded files take 10-30 seconds depending on length. The Gemini combined analyzer does deepfake detection, transcription, and scam classification in a single API call.

Protect the phone in your grandmother's hand

AI voice scams stole over a billion dollars from families globally in 2025. The voice is real. The crime is invisible. Verity makes it visible.