Konfir highlights certain patterns in a verification using alerts and insights. These are designed to make results easier to interpret - especially where information declared by the applicant differs from what Konfir can evidence from data sources.
Note: Alerts and insights are informational signals. How you use them (e.g. manual review, automated rules, or escalation) depends on your organisation’s policy and use case.
What they are: Alerts vs. Insights
Type | What it is | Why it exists |
Alerts | Rule-based flags that highlight mismatches, missing information, or higher-risk patterns. | Helps you spot areas that may need attention or explanation. |
Insights | Context signals that help explain a period or pattern (often neutral). | Adds context to support interpretation of the result. |
Important: Not every discrepancy will trigger an alert or insight. Flags appear only when their trigger conditions are met.
Where you’ll see them
Alerts and insights may appear in:
Konsole (on the timeline and within specific activities)
PDF exports (if enabled)
API results (returned in the verification payload, typically attached to relevant activities)
Types of Alert
Alert | What it means | When it may appear (trigger) | How to interpret |
Employment dates | Declared dates and evidenced dates do not align, or a gap/mismatch is detected. | When declared start/end dates differ beyond configured thresholds, or coverage suggests a gap. | Treat as a prompt to compare declared vs evidenced dates for that activity and decide whether the difference is material for your use case. |
Employer name | Declared employer name does not clearly match the evidenced employer name. | When the employer identifiers/names returned by sources do not map cleanly to the declared employer. | Treat as a naming/identity signal: confirm whether the mismatch is a legal entity/trading name/umbrella payer vs end client scenario. |
Company trading / validation | The employer appears inactive/not trading for the claimed period. | When employer validation checks indicate incorporation/dissolution timing doesn’t support the claimed dates. | Treat as a context flag about the employer’s status; interpret alongside employer naming structures and any supporting evidence returned. |
Undeclared employment | Evidence suggests an employment that wasn’t declared. | When sources show employment signals outside the declared timeline or for an employer not declared. | Treat as a completeness signal; whether it matters depends on what “complete history” means in your process. |
Suspicious pattern / fraud | A higher-attention pattern is detected. | When patterns associated with suspicious employers/role reuse are detected based on configured rules. | Treat as a higher-attention signal; interpretation and response should follow your organisation’s policy and control model. |
Types of Insight
Insight | What it means | When it may appear (trigger) | How to interpret |
Gap insight | Context signals detected during a gap period. | When evidence indicates activity/income patterns during a declared gap (often via banking). | Treat as explanatory context for a period (not proof of employment). Relevance depends on whether your use case cares about gaps vs income continuity. |
Alternative income | Non-standard or additional income patterns detected. | When income signals don’t fit a single-salary pattern or suggest additional pay sources. | Treat as context on income composition; how it’s used depends on whether you’re verifying employment history vs affordability/income. |
Trading dates | Incorporation/dissolution timing context for an employer. | When trading date evidence is available and relevant to an employment/activity. | Treat as supporting context for company trading questions and employer validation. |
Trading name mismatch | Name variation is detected across employer records. | When employer names differ across sources in a way that suggests trading/legal name differences. | Treat as context for employer name discrepancies; often resolved by mapping to the correct legal entity. |
Fraud | Fraud-related context surfaced in the result. | When fraud-related signals are present in the available evidence and ruleset. | Treat as context for higher-attention patterns; downstream handling should be policy-driven. |