Appearance
The Help Desk Console provides a centralized workspace for administrators and helpdesk agents to validate user identities, investigate suspicious requests, and securely carry out resolution workflows such as password resets and account unlocks.
Every sensitive help desk action is gated behind identity verification: imper.ai runs the same flow for each one - the user is verified first, and the action is only carried out once verification passes. A password reset is the most common example, but the same steps apply to any supported action.
The actions imper.ai can perform are:
Set password - set a new password for the user.
Unlock account - unlock the user's account.
Get a Temporary access pass (TAP) - issue a temporary access pass so the user can get into their account.
Additional actions can be enabled by request - contact imper.ai to add the actions your organization needs.
What You Can Do in the Help Desk Console
View all help desk-initiated requests
Review verification results, risk scores, action results, and detailed analysis reports
Send verification request flows through secure SMS or email links
Investigate unusual or failed verifications
Monitor verification analytics and trends, including success rates and risk patterns
How Help Desk Protection Works
Help desk protection follows the same steps for every gated action, from the initial request through to the completed action.
Step 1 - The User Requests an Action
The request may begin in one of the following ways:
The user reaches an automated helpdesk or IVR menu and selects an action such as a password reset or account unlock. See IVR Account Recovery.
The user initiates a request from an organizational portal.
The user contacts the helpdesk directly and speaks to a live agent (phone, chat, ticket).
Each request appears in the Requests Page Overview - the central view for helpdesk-driven identity validation - and is audited, so every request is logged and available for review.
Step 2 - A Verification Link Is Sent
The user is sent a secure Verification Portal link to their registered email or phone number. Depending on your organization's integration and configured verification method, the link can be sent automatically or by a live helpdesk agent; see Sending a Verification Portal Link for the agent flow. The link takes the user to the Self-Service Portal (SSP) to complete verification.
Step 3 - The User Completes Verification
The user opens the link and completes identity verification in the Self-Service Portal (SSP). Verification runs in two layers:
Infrastructure checks - silent, passive signals about the connection, location, device, and tooling, evaluated in the background with no interaction required. See Detectors and Risk Signals.
Security questions - if configured for your tenant, the user also answers dynamic, knowledge-based challenge questions generated from connected organizational systems and personal context. See Verification Questions Layer.
imper.ai scores the session and records the outcome (Passed, Failed, Timed Out, or Not Enough Questions), together with a risk score and a full analysis report; review the breakdown in the Analysis Report.
The result appears in the Verification tab of the request's drawer on the Requests Page Overview.
Step 4 - The Action Is Performed
Once verification passes, the action is completed in one of the following ways:
Self-service - the user performs the action themselves in the SSP (for example, resetting their own password or unlocking their account), with no agent involved.
Agent sends a link - an agent reviews the outcome and, if it passed, sends the user the relevant link using the Actions button in the Verification Summary panel.
Through your own policy - an agent completes the action using the organization's existing process, outside imper.ai.
If verification does not pass, no action is offered, and the agent follows internal escalation or remediation procedures (see Escalating Inconclusive Verifications below).
Escalating Inconclusive Verifications
When a verification comes back inconclusive - there isn't enough certainty to let the user proceed, but no clear indications of impersonation were found - imper.ai can apply an extra layer of confirmation before allowing a sensitive action.
The escalation is currently configured for the whole tenant (organization-wide). For a medium-risk (inconclusive) result, an admin sets the If medium risk detected - apply option to one of:
None - no escalation; the inconclusive result stands, so the user fails the verification and the action is not performed.
Verify with Document - the user confirms their identity with a government-issued document.
Verify with Manager - an agent confirms the user's identity through their manager by email.
Verifying with a Document
imper.ai can automatically add a document verification step. The user is asked to confirm their identity with a government-issued document, using either your organization's configured identity verification provider or the provider offered by imper.ai, and the outcome sets the final verification result.
For the full flow, see Verify with Document.
Verifying with a Manager
When a verification is inconclusive, imper.ai automatically emails the user's manager to confirm the user's identity. The manager approves or declines the request; on approval, the user can complete the action without further verification. High-risk failed verifications do not offer this option.
For the full agent and manager flow, see Verify with Manager.
Monitoring Help Desk Activity
The Helpdesk Dashboard gives insight into your helpdesk verification flows. It surfaces key metrics - completed verifications, successful actions in the Self-Service Portal, average daily verifications, and average verification duration - alongside charts that break verifications down over time, by risk score, and by final status (Passed, Failed, Timed Out, or Not Enough Questions), all for a selectable time range.
When imper.ai detects a suspicious event during verification - such as an abandoned verification or a user flagged as malicious - it is recorded as an incident on the Incidents Page, where security teams can investigate and track it to resolution.