About SlimeWatch
Our Mission
SlimeWatch empowers everyone with a phone to become a digital policeman — protecting their community without getting personally involved. Designed for personal and commercial use globally, our AI-powered platform creates a 20-mile protective radius anywhere in the world, connecting users to share verified incident reports and facial recognition alerts.
The platform learns and grows stronger with every user. Our AI continuously improves recognition accuracy and threat assessment from each verified incident. When businesses receive alerts about known offenders, they can deny access — building a reputation-based deterrent where repeated rejections bar troublemakers from services.
By combining advanced facial recognition with OpenAI's GPT-4o-mini Vision, we validate genuine incidents, categorize behavior, and assess threat levels — replacing the traditional policeman on the street with a global network of digital watchkeepers.
Who We Serve
SlimeWatch serves individuals and businesses globally who want to protect their communities and establishments from repeat offenders.
Personal Use Cases
- Community Members: Report antisocial behavior in your neighborhood without confrontation — be a digital watchkeeper from your phone.
- Parents & Residents: Document incidents near schools, parks, or residential areas to protect children and families.
- Public Transport Users: Record harassment or threatening behavior on buses, trains, or stations to warn others in your area.
- Anyone with a Phone: Become part of the global safety network — your reports help protect everyone within 20 miles.
Commercial Use Cases
- Retail & Shops: Document and prevent shoplifting, theft, and aggressive customer behavior with video evidence and facial recognition.
- Bars, Pubs & Clubs: Identify individuals involved in violence, rowdiness, or disruptive behavior to maintain safe environments for patrons.
- Restaurants & Cafes: Combat dine-and-dash incidents by documenting offenders and sharing alerts with nearby establishments in real-time.
- Hospitals & Healthcare: Protect staff and patients by documenting abusive or violent behavior in A&E departments and healthcare facilities.
How It Works
- Video-Only Upload: Users upload incident videos (only videos with detectable faces are accepted). Mobile videos automatically include GPS location and timestamps.
- AI Validation: OpenAI GPT-4o-mini Vision analyzes the video to verify legitimate antisocial behavior, categorize the incident, and assess urgency levels.
- Facial Recognition Processing: Advanced face detection extracts 512-dimension facial descriptors from video frames for accurate matching.
- Global Community Sharing: Validated reports are shared with all users within a 20-mile radius anywhere in the world.
- Real-Time Alerts & Access Control: When anyone scans faces using their camera, they receive instant alerts if a person matches known incidents. Businesses can use these alerts to deny access.
Technology
- InsightFace ArcFace — Server-side facial recognition with 512-dimension embeddings
- OpenAI GPT-4o-mini Vision — AI-powered video analysis and incident validation
- Cosine similarity matching — Accurate face matching with 0.6 threshold
- BFS grouping algorithm — Advanced face clustering for subject identification
- Secure cloud storage — Enterprise-grade video and data protection
Contact Information
Phone: 07368 488822
Email: trevor.smith@kanndoo.org
Visit our Contact Us page for more information or to send us a message.