πŸ† Conrad Challenge 2026 | Cyber-Technology & Security

Protecting Voices,
Stopping Scams

VocalGuard uses advanced AI to analyze phone conversations in real-time, detecting vishing attacks with 85%+ accuracy to protect vulnerable populations from the $39.5 billion annual fraud crisis.

85%+
Detection Accuracy
15s
Analysis Time
$39.5B
Annual Fraud Losses
🎯
Real-Time Analysis
Linguistic Threat
75%
Emotional Manipulation
62%
Voice Authenticity
48%
⚠️ HIGH RISK DETECTED
Scroll to explore

Our Story

Fighting fraud with innovation and compassion

πŸ’‘

The Problem

In 2023, phone scams stole $39.5 billion globally, with elderly populations losing $3.4 billion. Traditional solutions only block known numbers, leaving victims defenseless against sophisticated psychological manipulation and AI-generated voices. We witnessed our own families nearly fall victim to urgent "bank security" calls that bypassed every existing protection.

🎯

Our Mission

VocalGuard democratizes enterprise-grade fraud protection for everyone, especially vulnerable populations. We believe safety shouldn't depend on technical literacy or financial resources. Using cutting-edge AI, we analyze what scammers say, how they say it, and whether their voice is even realβ€”providing a shield against the invisible threat of vishing attacks.

πŸš€

Our Vision

A world where voice communication is trusted again. Where grandparents can answer their phones without fear. Where AI protects against AI-enabled fraud. VocalGuard is building the global standard for voice fraud detection, creating an ecosystem where scammers face insurmountable technical barriers, making vishing unprofitable and obsolete.

🀝

Accessibility First

Powerful protection for everyone, regardless of technical ability

πŸ”¬

Science-Driven

Evidence-based AI validated through rigorous testing

πŸ›‘οΈ

Privacy Protected

Zero audio storage, instant deletion after analysis

🌍

Global Impact

Protecting vulnerable communities worldwide

How VocalGuard Works

Multi-modal AI detecting fraud invisible to humans

1
🎀

Upload Audio

Record or upload suspicious call

β†’
2
πŸ€–

AI Analysis

3 simultaneous detection layers

β†’
3
πŸ“Š

Risk Score

Real-time threat assessment

β†’
4
πŸ›‘οΈ

Protection

Actionable security alerts

πŸ“

Linguistic Threat Score

50% Weight

Analyzes conversation content using Whisper AI transcription and sentence-transformers. Detects 47 scam-specific keywords and phrases like "verify OTP," "account suspended," "urgent action required."

Natural Language Processing Semantic Analysis Pattern Recognition
🎭

Emotional Manipulation Score

30% Weight

Measures vocal stress patterns, pressure tactics, and urgency indicators. Detects pitch variance, energy fluctuations, and speaking rate changes that reveal psychological manipulation attempts.

Acoustic Analysis Pitch Detection Energy Mapping
πŸ”Š

Voice Authenticity Score

20% Weight

Identifies AI-generated voices and deepfakes with 87% accuracy. Analyzes spectral flatness, harmonic ratios, and distinguishes synthetic audio from natural human speech patterns.

Deepfake Detection Spectral Analysis Harmonic Verification

Real Results

85%
True Positive Rate
12%
False Positive Rate
12.3s
Average Processing
1,200+
Verified Scam Dataset

Meet The Newtonians

Students building the future of voice security

Team ID 11601
Chapter Global
Category Cyber-Technology & Security
Institution St. Joseph Higher Secondary School
πŸ‘¨β€πŸ’»

Tawhid Bin Omar

Co-Founder & Lead Developer

Full-stack developer passionate about AI/ML and cybersecurity. Built the complete VocalGuard platform from backend architecture to AI model integration. Specializes in cloud infrastructure and real-time processing systems.

Python AI/ML FastAPI Firebase
πŸ“§ tawhidbinomar@gmail.com
πŸ‘¨β€πŸ”¬

Tapantor Das

Co-Founder & Research Lead

AI researcher focused on audio processing and fraud detection algorithms. Developed the multi-modal risk scoring system and conducted validation studies. Drives innovation in voice authenticity detection.

Machine Learning Audio Processing Data Science Research
πŸ“§ tapantordas2008@gmail.com

Our Journey

October 2025

Concept & Research

Identified vishing as critical gap in cybersecurity landscape

November 2025

MVP Development

Built functional prototype with $0 budget using open-source tools

December 2025

Beta Testing

47 users tested system, achieved 85% accuracy on 340+ calls

January 2026

Conrad Challenge

Competing for innovation in Cyber-Technology & Security

Our Impact

Protecting vulnerable populations, preventing billions in losses

πŸ‘΄

Elderly Protection

73% of our beta users were 60+, the demographic losing $3.4B annually to phone scams. VocalGuard provides simple, accessible protection for those most vulnerable to psychological manipulation tactics.

πŸ’°

Financial Security

Projected to prevent $25M in fraud losses annually with just 10,000 users. Every detected scam saves victims an average of $2,500 and prevents devastating financial and emotional damage.

🌍

Global Accessibility

Freemium model ensures powerful protection reaches everyone, not just those who can afford premium security. 100 free monthly analyses democratize enterprise-grade fraud detection.

πŸ”¬

Industry Innovation

First consumer-grade solution combining linguistic, emotional, and voice authenticity analysis. Setting new standards for voice fraud detection in the emerging AI deepfake era.

"VocalGuard detected a scam that my bank missed. The urgency in the caller's voice had me reaching for my card, but the 75% threat score made me pause and verify. It saved me thousands."
Beta User, Age 68 Retired Teacher

Ready to Protect Yourself?

Join the fight against vishing. Try VocalGuard free today.

Launch VocalGuard