EdgeTrace

EdgeTrace unifies your entire camera network into a single, powerful intelligence platform–helping public safety teams detect threats faster, respond in real time, and protect the community.

Founded by engineers from Cruise and xAI, we're a Y Combinator-backed team building AI that works in the chaos of real-world public safety operations. We move fast, value impact over process, and measure our success by the crimes our technology helps solve.

If you want to build mission-critical technology that directly protects communities, let's talk.

Frontend/Backend: TypeScript everywhere with Next.js + NestJS, because type safety matters when bugs could mean missing suspects AI Pipeline: Python for rapid model development, C++ (Userver framework) for production inference where every millisecond counts GPU Optimization: Custom CUDA kernels and TensorRT optimization to search thousands of video streams simultaneously Message Architecture: RabbitMQ orchestrating distributed video processing across GPU clusters

The Technical Challenges: Building computer vision models that work on grainy night footage, rain-covered lenses, and 20-year-old cameras Maintaining sub-second search latency across petabytes of video data Creating legally-defensible audit trails for every search and detection Designing systems that never go down - because crime doesn't stop for deployments