How Identiguard works
Every feature in Identiguard is built on the same flow. If you can explain this page, you can explain the product.
1. Events come in
Section titled “1. Events come in”Each identity platform sends events in its own way.
| Platform | Method | What that means for the client |
|---|---|---|
| Okta | An event hook that Identiguard creates in the client’s Okta org | Events arrive within seconds. Directory agents (AD and LDAP) are checked on a timer. |
| Auth0 | A log stream the client creates, pointed at an Identiguard endpoint | Events arrive within seconds. Identiguard holds no Auth0 credentials. |
| Entra ID | Identiguard polls Microsoft Graph on an interval you set. The default is 5 minutes. | Events arrive on the poll interval, not instantly. |
All access is read-only, apart from the one event hook Identiguard creates in Okta.
2. Events are graded
Section titled “2. Events are graded”Each event gets a severity: DEBUG, INFO, WARN, ERROR or CRITICAL. Identiguard uses the provider’s own severity first and falls back to its own baseline. The feed shows this as a risk rating from 1 to 5. See Severity and risk rating.
A client can add their own importance for the events they care about with event priorities. Priorities change how events are flagged and sorted in the feed. They do not change alerts or incident scores.
3. Events appear in the live feed
Section titled “3. Events appear in the live feed”The dashboard shows the feed in real time. From any event an analyst can open the details, ask Alex about it or dismiss it. With ServiceNow connected, they can also raise a ticket.
4. Detectors open incidents
Section titled “4. Detectors open incidents”Detectors watch for patterns across events. When one fires, Identiguard opens or adds to an incident, scores it, and lays the events out as an attack story. Identiguard also learns what is normal for each identity and flags behaviour outside that baseline.
5. Alerts and tickets go out
Section titled “5. Alerts and tickets go out”Events at or above a chosen severity are forwarded to Slack, Teams or email. An analyst can turn any event into a ServiceNow incident.
6. Data is kept, then removed
Section titled “6. Data is kept, then removed”Raw events are kept for 48 hours by default, and up to 30 days depending on the plan. Summaries used for reports are kept for up to 13 months. Reports are removed 180 days after they were created. See Security and data.
7. Reports summarise the period
Section titled “7. Reports summarise the period”A health check report reads the tenant’s data for the reporting window and produces a PDF with findings.