Have questions about setting up and using Revelation AI? You’re in the right place. This FAQ covers the most common topics admins and users ask about — from security and access controls, to supported models, cost management, and day-to-day usage tips. Use this section as a quick reference whenever you need clarity on how Revelation AI works, what it can do, and how to get the most out of its features.

For issue troubleshooting, please refer to the Common Issues and Fixes list.

Is Revelation AI secure?

Yes. Revelation AI is built with strict tenant isolation and role-based access control. AI suggestions are generated within your organization’s context and never shared across tenants.

Who can use AI Assist?

Only users granted access via the Admin Settings tab. Super Admins can enable or disable AI Assist per team or access level.

What models does Revelation AI support?

Currently, Revelation AI supports Azure OpenAI GPT models using the Global Standard deployment type. The recommended default is gpt-4.1-mini, optimized for cost and performance. These models may change over time, so it's important to review the current options and their capabilities before deploying. Super admins can deploy and benchmark new models as they become available, and only models available in the selected region will be listed.

Can I edit AI-generated responses?

Absolutely. AI Assist suggestions are editable before sending. We encourage users to review and personalize replies to match your team’s tone and style.

How do we control costs?

Admins can set Azure budgets and monitor token usage via the Cost Management dashboard in the Azure AI Foundry portal. Revelation AI also supports throughput limits (RPM/TPM) to manage performance and spend.

Can I create resources directly from the Admin UI?

Yes - the signed-in Azure user must have the Cognitive Services Contributor (AIM) role assignment on the selected Subscription. We recommend pre-creating a resource group and subscription in the Azure Portal. If you want to create the resource group in the Revelation admin area, the user will need the (privileged) Contributor role assigned on the subscription.

What if I need to rotate an API key?

Use the “API Key” action in the Deployments tab to update your Azure OpenAI key. This supports key rotation and quick replacement in case of exposure.

Can I test different models?

Yes. Revelation AI supports multiple deployments. Use the “Add New” button in the Deployments tab to test and compare models. Only one can be set as the system default at a time.