AI Cloud Provider Integration

AI Integration describes how 12Port integrates with a customer-managed AI cloud provider to process event data for AI-powered analysis. This integration enables 12Port to securely transmit eligible report data and selected prompts to the configured AI service, where analysis and summary generation are performed in accordance with the customer’s cloud environment, security controls, and usage policies.

This section outlines the supported integration model, configuration requirements, and operational considerations for enabling AI-powered insights using a customer-provided AI platform.

Note

The Intelligent Tagging, Session Intelligence, and Analytical Models AI/ML features operate using local processing only and do not require integration with a Cloud AI provider or transmission of data to external AI services.

How to Integrate

How to integrate 12Port with your AI Cloud Provider:

1. Login to your 12Port tenant with an Administrator or AI Manager account.
2. Navigate to AI > Integrations and click the Add button.
3. Enter a unique, but recognizable Name for this integration.
4. Select your AI Cloud Provider from the dropdown.
5. Enter your AI Cloud Provider Enterprise API Key. This API Key is generated through your AI Cloud Provider. Note that 12Port does not provide an API Key.
6. Choose the AI Cloud Provider Model to be used for analysis. Note that each model will have different costs and usage or token limits, choose the model that will be best for your intended use cases. We recommend experimenting with -mini or -nano variants first.
7. Toggle on the Enabled parameter.
8. Use the Test Connection button to confirm authentication and connectivity to the AI Cloud Provider.
9. Click Save to save the integration.

Tip

For operational flexibility and cost optimization, 12Port supports the creation of multiple integrations with the same AI provider, each configured to use a different AI model. This allows administrators to align model selection with the processing requirements of specific prompts and report types.

For example, reports that can be accurately analyzed using lightweight models (such as -mini or -nano variants) can be routed through an integration configured for those models, providing faster response times and lower operational cost. More complex analyses that require advanced reasoning or deeper contextual processing can be executed using a separate integration (with the same API Key) configured with a larger, full-capability model, which may incur higher cost and longer processing time.