Enrichment modes

Asset Intelligence supports three enrichment modes that control the balance between enrichment coverage and accuracy.

Note:
Enrichment mode configuration is available only in Vantage. Guardian and CMC appliances use the mode configured in Vantage for their enrichment results. For more details, see Product availability.

The enrichment mode determines how Asset Intelligence (AI) handles uncertain predictions. Configure the enrichment mode at the organization level.

Available modes

AI offers the following enrichment modes:

High confidence only
The most conservative mode. AI only applies enrichments with very high confidence scores. This mode minimizes false positives but results in lower enrichment coverage. Use this mode in environments where accuracy is critical.
Safe enrichment (default)
A balanced mode that discards uncertain predictions. This mode provides a good balance between accuracy and coverage. Safe enrichment is the recommended mode for most deployments.
Aggressive enrichment
The most permissive mode. AI maximizes enrichment coverage by including uncertain predictions. This mode has a higher false positive rate. Use this mode when broad asset discovery and classification is the priority.
Figure 1. Enrichment mode configuration in Vantage
Screenshot of the Vantage UI showing the enrichment mode configuration s etting with the three available options: High confidence only, Safe enrichment, and Aggressive enrichment

Choose the right mode

Consider the following factors when selecting an enrichment mode:

  • Environment sensitivity: Critical infrastructure environments benefit from high confidence only mode to minimize incorrect enrichments
  • Asset visibility goals: If broad coverage is the priority, aggressive enrichment provides the widest visibility
  • Feedback loop: Use the enrichment feedback mechanism to report incorrect enrichments and improve future accuracy regardless of the mode

Impact of mode changes

Changing the enrichment mode affects only future enrichment requests. Existing enrichments are not retroactively modified. After changing the mode, new enrichment results reflect the updated confidence thresholds on the next enrichment cycle.