Student Team
Sanshudh Jain
KAVACH AI Team
Kaashvi Siotia
KAVACH AI Team
Adamya Pratap Singh
KAVACH AI Team
India’s intelligent shield against examination fraud: a multi-layer, AI-native security ecosystem that protects merit and preserves public trust.
Pre-exam theft from printing presses, storage facilities, and insider channels.
Impersonators with forged documents sitting exams on behalf of real candidates.
Bluetooth earpieces, smartwatches, and concealed electronic aids.
Telegram channels, WhatsApp groups, and corrupt examination centre collusion.
Existing security measures are reactive, manual, and easily bypassed by coordinated fraud networks equipped with modern tools.
Fraud is detected after it has already occurred.
Manual processes are vulnerable to fatigue, bias, and corruption.
Current systems cannot identify high-risk centres or suspicious patterns before exam day.
Modern devices and proxy techniques routinely defeat basic screening.
Dynamic, encrypted, last-minute paper assembly with regional randomisation.
Face recognition, liveness detection, Aadhaar integration, and behavioural biometrics.
AI CCTV analytics, Bluetooth signal detection, and crowd behaviour analysis.
Predicts high-risk leak scenarios using historical fraud data and centre-level indicators.
Creates tamper-proof logs from paper creation to result declaration.
Updates dynamic risk scores and routes alerts to command teams.
AI creates structurally equivalent but uniquely sequenced paper versions, transmitted securely with tamper-alert triggers.
Multi-modal biometric fusion verifies candidates in under 30 seconds and makes proxy success nearly impossible.
Computer vision and RF sensors identify suspicious movements, hidden devices, and coordinated cheating signals in real time.
Uses historical fraud clusters, communication signals, access logs, and behavioural anomalies to predict risk before exam day.
Records paper generation, access, candidate verification, invigilator actions, and result processing in auditable logs.
Highlights suspicious candidates and centres with confidence scores for rapid invigilator and central command action.
Dynamic paper generation eliminates single-point failure.
Multi-modal biometrics reduce proxy candidate success.
Real-time AI detection disrupts coordinated fraud networks.
Blockchain audit trails improve transparency and accountability.