Data Source Health Monitoring¶

AI Analyst automatically monitors your connected data sources every day. If a connection fails or query performance degrades significantly, AI Analyst sends an email alert to your organization's Admins so you can fix issues before they affect your team's analysis.

What gets monitored¶

AI Analyst runs two types of checks on each data source daily:

Connection failure alerts¶

AI Analyst tests whether each data source is reachable. If a connection fails, all Admins with this alert enabled receive an email notification.

Key behaviours:

• The alert is sent once per outage — you won't receive repeated emails for the same ongoing failure
• When the connection recovers, the alert resets and will fire again if a future failure occurs
• The email includes a direct link to the affected data source in AI Analyst Studio

High latency alerts¶

AI Analyst measures the average query duration across your data source over the past 7 days. If the average exceeds 30 seconds, an alert is sent.

Key behaviours:

• Requires at least 5 completed queries in the past 7 days before alerting (to avoid false positives on rarely-used sources)
• A 14-day cooldown applies per data source — if latency remains high, you'll receive a follow-up alert after 14 days
• The alert is sent to all Admins at the same time; the cooldown is shared across your organization

Managing your notification settings¶

You can enable or disable each alert type individually. Settings are per-user, so each Admin can choose which alerts they want to receive.

1. In AI Analyst Studio, go to Settings → Notifications
2. Toggle Connection failure alerts and/or High latency alerts on or off

Both alerts are enabled by default for all Admins.

Who receives alerts¶

All Admins can visit Settings → Notifications to manage their preferences

Each Admin's settings are independent — one Admin can opt out of high latency alerts while others remain subscribed.