Manage AI models: see which models are available to your account, track versions of your own models, and convert uploaded models for edge deployment. See Upload your model.
Deploy models to edge devices: assign or swap model pipelines on any device remotely. Group devices and push a pipeline change to all of them at once.
The diagram below shows how model sources, the cloud platform, and edge devices relate.
Nx AI platform — model sources are converted and stored in the cloud; the edge device runs pre-processing, inference, and post-processing on live input
Key concepts
Understanding the following terms helps explain how models, devices, and access control relate to each other in the Nx AI Cloud platform:
Site: An installation of one or more Nx Servers registered to your cloud account. Sites appear in the Sites tab and can be online or offline.
Server: An individual Nx Mediaserver process running inside a site. A site can have multiple servers; each server hosts a set of devices.
Device: A camera or other input source connected to a server. AI pipelines are assigned and run per device.
Organization: A grouping of sites and users for access control. Models and catalogues must be connected to an organization for the devices in that organization to see them.
Catalogue: A named collection of AI models. Catalogues are assigned to organizations to control which models are available to which sites.
Pipeline: One or more chained AI models assigned to a specific device, defining what runs on that device's video stream.
If a custom model is not visible when assigning it to a device, the most common cause is that the model's catalogue is not connected to the organization that owns the site. See Upload your model for details on setting this up.
Accessing the Nx AI Cloud interface
The Nx AI Cloud admin interface is a web application. After logging in with your Nx account, you can manage your models and devices from any browser.
The Nx AI Cloud admin interface — log in with your Nx account to continue
