EyePop.ai

EyePop.ai is a visual intelligence platform that enables users to train and deploy custom AI models to detect people, objects, actions, and conditions in images and video. It solves the problem of manually reviewing large volumes of visual content by automating the extraction of structured, searchable information from live or recorded footage. The platform runs in the cloud or fully on-premise on edge devices, allowing organizations to process data where it lives for latency, privacy, or policy reasons.

The platform is built around composable "Abilities"—pre-built or custom visual tasks that can be chained into pipelines. Key capabilities include self-service training, auto-labeling, and structured output formatting. It integrates with hardware such as Qualcomm Snapdragon Dragonwing, NVIDIA Jetson ORIN (NX Nano, AGX, Thor), and offers SDKs and APIs for Python and Node. The platform also provides modules for ingest-time intelligence, edge inference, and VMS enrichment, with versioned abilities and observability tools.

EyePop.ai is designed for operations teams in surveillance, marketplaces, broadcast media, and CDNs, and does not require in-house machine learning expertise. It offers a cloud-based Production plan at $200 per month and an Enterprise cloud tier starting at $800 per month, both with usage-based compute units. On-premise deployment is priced per device, starting at $250 per box per month, with a one-time setup fee beginning at $5,000 and a free three-month license for one internal lab box for testing.