Skip to main content

​
Introduction

Edge Inference Services is not yet available in production. The Inference tab currently shows a Coming Soon placeholder. This article describes the intended functionality.
Edge Inference Services is integrated into the ZEDEDA console and accessible directly from the Inference tab. This article describes the layout of the Inference tab and how to move between its sections.

​
Prerequisites

  • An active ZEDEDA user account with access to the Inference tab.
  • A role that includes Inference permissions. Contact your ZEDEDA administrator if the Inference tab is not visible.

​
Open the Inference Tab

  1. Log in to your ZEDEDA console at your enterprise URL (for example, https://zedcontrol.zededa.net).
  2. Select Inference in the bottom navigation bar.
  3. The Inference tab opens, showing the left sidebar with the available sections.
The left sidebar contains top-level pages and three collapsible section labels. Select a section label (Build, Deploy, or Monitor) to expand or collapse the items beneath it.

​
Dashboard

Dashboard is the default landing page when you open the Inference tab. It provides an overview of your inference environment and is separate from the Infrastructure Dashboard. See Inference Dashboard. Locations also appears in the Inference sidebar. It is a shared page available in both service areas. See Locations.

​
Build section

Select Build to expand or collapse this group. It contains the following item. Build with ZEIA opens the full-screen ZEIA workspace. ZEIA is the AI assistant built into the console. It answers questions in plain language using your live fleet data, EVE-OS (Edge Virtualization Engine) telemetry, and ZEDEDA documentation. You can also open ZEIA from the terminal bar at the bottom of any page.
ZEIA is not yet available in production. The ZEIA panel currently shows a Coming Soon placeholder.
See Use ZEIA to Query and Operate Your Fleet for ZEIA procedures.

​
Deploy section

Select Deploy to expand or collapse this group. It contains the following items. Models shows your ML model inventory. The Models list displays each model with its name, description, ML framework, organization, tags, and last-updated date. From this view you can upload models, import from external providers, filter by framework or organization, and open individual model detail pages. See Models for full model management procedures. Providers is where you configure connections to external model sources such as HuggingFace, MLflow servers, and S3-compatible storage. After you configure a provider, you can import models from it directly into your model inventory. See Providers for provider configuration procedures. Evaluation is where you benchmark models on real edge hardware before deploying them to production. The Evaluation view has two sub-views: benchmark runs (latency, throughput, and resource usage by device) and Device Pool (manage the set of devices available for benchmarking). See Inference Model Benchmarking for benchmarking procedures. Deployments is where you review and manage active model deployments across your edge fleet. Each deployment record shows the model, target cluster, deployment status, and configuration details. See Deployments and Validation for deployment procedures.

​
Monitor section

Select Monitor to expand or collapse this group. It contains the following items. Jobs shows the status of model import, benchmark, and deployment jobs. See Jobs and Monitoring for monitoring procedures.

​
Access Administration Settings

Manage organizations, users, and roles for Edge Inference Services through the Users and Roles icon in the header.
  1. Select the Users and Roles icon (a person silhouette with a small gear) in the header of the Inference tab. Hover over the icon to confirm the tooltip reads “Users and roles”.
  2. Select one of the following:
    • Organizations: Create and manage model organizations.
    • Users: Add, update, and remove users.
    • Roles: View available roles and their permissions.
See Organizations, Users, and Roles for administration procedures.

​
Update Infrastructure Services API Credentials

Edge Inference Services uses your Edge Infrastructure Services API credentials to communicate with the ZEDEDA platform, prepopulate deployment fields, and deploy models to ZEDEDA Edge Kubernetes Service clusters. If your credentials expire, update them through your user profile.
  1. Select your user avatar (the circular button showing your initials) in the top-right corner of the navigation bar.
  2. Select Profile Settings from the dropdown menu.
  3. On the Profile page, select the Infrastructure tab.
  4. Under Infrastructure Services API Credentials, select the Edit icon. If no credentials have been configured yet, select Add Credentials instead.
  5. Update the Infrastructure Services API Token and Infrastructure Services Host fields with your current credentials. The token is required and must be at least 10 characters. The host is optional.
  6. Select Save.
To obtain a new session token, see Authentication.

​
In This Section