Register & validate
Provide model metadata, expected inputs, output classes and runtime requirements before deployment.
Bring your own computer vision models into the same deployment, event, and analytics workflows as built-in intelligence.
Register and upload a model, validate runtime compatibility, assign cameras, configure regions of interest, and define event rules. Use common dashboards, APIs and notifications to deliver the results.
Provide model metadata, expected inputs, output classes and runtime requirements before deployment.
Choose cameras, devices, execution modes and regions of interest for each operational environment.
Map inference outputs to structured events, custom analytics, alerts and enterprise integrations.
Upload → Validate → Select cameras → Configure ROI → Run inference → Create events → Deliver analytics, API results and notifications.
TensorRT · NVIDIA GPU
ONNX Runtime CUDA
ONNX Runtime CPU
Illustrative product configuration
ILLUSTRATIVE IMAGESet each direction to match your camera view. Changing IN leaves OUT unchanged.
IN: right · OUT: left
Bring your infrastructure, services and operational response into one view. Start an ORVEXION conversation around your environment.