AI Insights FAQs

You need:

• A valid QTDMP (Quest Trusted Data Management Platform) account.
• An active Foglight license with AI access.
• At least one monitored database or infrastructure target.
• An internet connection (AI requires online access).

No. Foglight AI Insights requires an active internet connection and an online license.

• You can have a maximum of 2 active chats at a time
• AI outputs should be reviewed and validated by you before acting on recommendations
• Internet connectivity is required at all times
• Rate limits apply to API requests

About Data Accuracy:

Foglight AI Insights generates its answers based on the data stored in Foglight, so the underlying information it draws from is accurate. However, the AI’s interpretation of your question, or the way a question is phrased, can affect how focused or relevant the answer is. If a response seems off-target or too general, try rephrasing or refining your question with more specific details to get a more accurate result.

About Your Responsibility:

As with any AI-generated output, you remain responsible for reviewing and validating the results before implementing any changes.

No. Your data is not used for training AI models. Your queries and data remain private.

No. Your data is isolated and visible only to you based on your organization’s access controls.

All data processing takes place in U.S. regions within Quest’s Microsoft Azure environments.

Your data is protected using:

• Encryption in transit (TLS 1.2+)
• Encryption at rest (256-bit AES)
• Secure authentication (Microsoft Entra ID)
• Strict access controls based on your role

Yes, the following data is stored:

• Chat Conversations: Your chat history and AI responses are stored only in your browser’s local storage, under your control. This data is not encrypted.
• MCP Server Configuration: MCP server configurations and settings are stored to maintain your setup and enable consistent query shortcuts.

All data stored in Quest repositories is encrypted and protected according to our security policies. Your data remains private and visible only based on your organization’s access controls.

Not necessarily. MCP describes how an AI client calls tools. Where the AI model runs and what data is shared depends on your deployment and the client configuration.

Usually no. Prefer the higher-level tools (Alarms, PI, TopSQL) for common operations. Use GraphQL for advanced/custom data retrieval.

That is expected in some workflows. Use the returned IDs to call the next tool in the chain. Use the decompress tool when a response provides compressed/encoded SQL text.