Sherlock

Visual Preview

Website thumbnail

Sherlock is an advanced interview integrity platform designed to detect AI-assisted cheating during remote interviews. The system uses a multi-layered approach combining vision, hearing, and reasoning technologies to monitor candidate behavior. Sherlock's technology includes device monitoring (microphone, camera, and screen), audio detection for identifying external assistance patterns, and visual analysis with 80-85% accuracy for detecting off-screen reading or inconsistent thinking patterns. The platform integrates with popular interview platforms like Zoom, Teams, and Meet through API or browser extensions, requiring only standard webcams and microphones from candidates. Sherlock maintains a false positive rate below 0.3% and provides real-time alerts during interviews followed by comprehensive reports with timestamps and confidence scores. The system is GDPR compliant, encrypts all data, and stores only metadata relevant to cheating detection. Sherlock is effective for both technical and non-technical interviews, detecting various cheating methods including AI tool usage, off-camera assistance, pre-written answers, and screen-sharing setups. Implementation typically takes 1-2 business days with comprehensive onboarding support available.