AI/ML Service Activation for Self-Managed Environments

Enable JFrog AI/ML frontend service to integrate ML model building and deployment with enterprise security controls.

JFrog AI/ML integrates everything machine learning and data science teams need to build and scale ML models in production, yielding a unified AI solution for every need. AI/ML capabilities also include access and use of AI Catalog, including shadow AI, MCP registry, skills registry and models catalog, and so on.

📘
Note:
By default, AI/ML capabilities are disabled for Self-Managed JFrog subscriptions. If you want to use JFrog's AI and machine learning capabilities, you need to follow the instructions in this topic and then the instructions in Install the JFrogML Bridge.

AI/ML teams can reliably and rapidly move models from experimentation to production, while applying proven development security and control practices across the lifecycle.

For Self-managed JFrog subscriptions, JFrog AI/ML uses a frontend/backend model.

The backend is hosted by JFrog in the cloud.
The frontend service runs on your JPD. It is included in the Artifactory installer and must be enabled.

📘
Note
Do not enable the JFrog AI/ML service on Artifactory Edge nodes.

Prerequisites

Before enabling JFrog AI/ML, verify that your environment meets the following requirements.

Requirement	Details
Network connectivity	Direct internet connection to `https://bridge.ml.frog.io` (proxy not supported)
JFrog platform	Self-managed JPD with Artifactory 7.x
License	Enterprise+ (contact JFrog to confirm AI/ML entitlement)
Edge nodes	Do not enable on Artifactory Edge nodes
Version	All dependencies in the bridge have 'matching' versions
Entitlement	AI Catalog entitlement

Enabling JFrog AI/ML

Verify back-end connectivity: your JFrog Platform instance must reach https://bridge.ml.frog.io via a direct internet connection. Proxy configuration is not supported.
Enable the JFrog AI/ML front-end service using one of the following methods:
- Helm installation: The Artifactory Helm chart supports the ml.enabled and jfconnect.enabled keys. Set both when you install or upgrade the chart:
```
--set ml.enabled=true 
--set jfconnect.enabled=true
```
- Manual configuration: Go to the $JFROG_HOME/artifactory/var/etc directory. Edit the system.yaml file as follows:
```
frontend:  
  ml:    
    jfrogMlAppState: VISIBLE
jfconnect:
  enabled: true
```

Restart Artifactory.
Verify that the JFrog AI/ML service is running. See Monitoring Service Statuses.

In the JFrog Platform interface, click Platform module > AI/ML and follow the instructions to install the essential components for operating AI/ML in your self-managed environment.
In order to start using the AI/ML capabilities, you must follow the instructions in the Install the JFrog ML Bridge section in the AI Catalog in Self-Managed Systems topic.

Frequently Asked Questions

This section provides answers to frequently asked questions.

plusFAQs

Q: Is a direct internet connection required for JFrog AI/ML?

A: Yes. The JFrog AI/ML front-end service on your JFrog Platform instance must connect directly to https://bridge.ml.jfrog.io (the JFrog-hosted back-end). Proxy configurations are not supported for this connection.

Q: Which license tier includes JFrog AI/ML?

A: JFrog AI/ML is available on Enterprise+ self-managed installations. Contact JFrog to confirm AI/ML entitlement for your license. The feature is not available on Edge nodes.

Q: How do I verify that JFrog AI/ML is running after enabling it?

A: After restarting Artifactory, navigate to Platform module > AI/ML in the JFrog Platform UI. Follow the in-UI instructions to complete the setup. You can also monitor the service status via Monitoring Service Statuses.

Prerequisites

Enabling JFrog AI/ML

Frequently Asked Questions

Q: Is a direct internet connection required for JFrog AI/ML?

Q: Which license tier includes JFrog AI/ML?

Q: How do I verify that JFrog AI/ML is running after enabling it?

Related Topics