Building observability and robustness systems for intelligent infrastructure.
- Working on AI safety, perception robustness, and realtime observability.
- Building SIGNALWATCH: public AI observability for real source monitoring, evaluation visibility, and safety-critical perception.
- Interested in deployment monitoring, temporal robustness, perception failures, and model behavior under real-world uncertainty.
- Systems stack: TypeScript / Next.js, Python / FastAPI, WebSockets, Computer Vision, Docker, Railway, Vercel.
jgsops.dev | signalwatch | barpath-ai | Squat-Judge
Reality is the final benchmark.