Sentinel is an open-source AI video intelligence platform built for defense, critical infrastructure, and high-security environments. It enables real-time multi-camera ingestion, YOLO-powered object detection, behavioral analysis, and natural language search across footage. Designed for secure and air-gapped deployment, Sentinel provides a unified command interface for mission-critical surveillance systems.
No reviews yetBe the first to leave a review for Sentinel AI Video Intelligence
Maker
📌
Hi everyone — Arsh here, founder of ABG.
Sentinel was built from a simple observation: most surveillance systems are feature-heavy but system-light. They perform well in controlled demos, yet struggle in real deployment where scale, operator workload, infrastructure variability, and security constraints matter more than model accuracy.
We wanted to approach video intelligence as architecture — not just detection.
Instead of building another AI overlay on CCTV, we focused on creating a unified system that supports real-time multi-camera ingestion, structured detection pipelines, natural language search, alert orchestration, and secure deployment — including air-gapped environments.
During development, our thinking evolved from “add more models” to “design better systems.” We spent more time on workflow, structure, and deployment design than on adding flashy capabilities. The result is Sentinel Core — now open-sourced — forming the foundation for a broader, deployment-ready platform.
We’re sharing this early to invite feedback from engineers, operators, and builders working on security and infrastructure systems.
Happy to answer any technical questions.