Traces
Traces AI provides an AI-powered video analytics platform that transforms existing CCTV infrastructure into intelligent security assets. The platform reduces false alarms by over 95% using generative AI, enables forensic search across 2,000+ appearance parameters beyond facial recognition, and supports natural language video search. It is hardware-agnostic, cloud-scalable from 30 to 30,000 cameras, and available as an API.
- Artificial Intelligence
- Cybersecurity
- Data & Analytics
- Software Only
Funding
Funding not disclosed
Founders
Founder details are not available yet.
Product
Problem
Traditional CCTV systems generate high volumes of false alarms that overwhelm security teams, leading to alarm fatigue and delayed response to genuine threats. Investigating incidents across thousands of hours of video footage is time-consuming and often fails when facial recognition is impossible due to obscured or low-quality imagery.
Solution
Traces provides an AI-powered video analytics platform that works with existing CCTV infrastructure to deliver real-time threat detection and forensic investigation capabilities. The platform uses generative AI to filter false alarms, reducing them by over 95%, and provides natural language video search that lets users identify relevant footage with text prompts. Its forensic search technology analyzes over 2,000 appearance parameters—including clothing, body shape, hairstyle, and items in possession—to identify and track individuals across multiple cameras even when faces are not visible. The platform is hardware-agnostic, cloud-scalable from 30 to 30,000 cameras, and available as an API for integration into existing security operations.
Target Audience
Primary customers are security operations centers, retail chains, transportation hubs, and property management companies that need to reduce false alarms, accelerate investigations, and derive business intelligence from existing video surveillance infrastructure.
Features
- Generative AI false alarm filtering that processes over 1 million alarms weekly and reduces false alarms by over 95%
- Forensic search with over 2,000 appearance parameters for person identification and tracking beyond facial recognition
- Natural language video search allowing users to input text prompts to locate relevant surveillance footage
- Hardware-agnostic deployment compatible with any existing CCTV cameras without replacement
- Cloud and on-premise deployment options with sub-second alarm verification latency and 99.9% uptime SLA
- GDPR/CCPA compliant with local data residency, no personal data storage, and optional face blurring
- AI Assistant enabling natural language camera operations and investigation workflows
- Multi-cloud infrastructure distributed across AWS and GCP with AES 256 encryption at rest and TLS 1.2 encryption in transit