Cloud-based video AI platform for residential security and safety monitoring
InstaVision operates a cloud-based video AI platform that offloads computer-vision inference from edge devices to AWS, MongoDB, and PostgreSQL infrastructure—a cost-optimization pattern that lets them serve residential customers without requiring expensive on-device AI chips. The tech stack (Python, PyTorch, TensorFlow, OpenCV) and project pipeline (intruder detection, face recognition, baby health monitoring) reflect a multi-domain safety focus. Hiring velocity is accelerating with 5 engineering roles open, all posted in the last 30 days, though internal friction around product requirements and development process suggests they're scaling faster than org discipline.
InstaVision builds a cloud-based video AI platform for residential security and safety, launched in 2023 from Carlsbad, California. The platform uses computer-vision models deployed on AWS infrastructure to analyze IP camera feeds and detect threats, occupants, and environmental hazards—breathing patterns, fire, intruders, vehicles, animals, packages, gestures, and faces. By centralizing AI inference to the cloud rather than embedding it on-device, the company reduces hardware costs and expands addressable market. The product surface spans web (React, TypeScript, Vite) and mobile (Android, iOS), with backend services running Python and Go. The 11–50-person team is engineering-focused, with early-stage hiring concentrated in India.
Python, Go, PyTorch, TensorFlow, OpenCV for AI; AWS, MongoDB, PostgreSQL, Cassandra for backend; React, TypeScript, Vite for web frontend; Docker, Kubernetes for infrastructure. Payments via Stripe.
Intruder detection, face recognition, baby health monitoring, and scalable web frontend applications for their cloud-based video AI platform.
Other companies in the same industry, closest in size