Search Guide
Use the Search tab for hybrid (semantic + structured) triage across the case corpus.
Filters
Applied filters appear as badges above the results area. Available filter categories include:
Source — filter by data source
Taxonomy — filter by fraud classification (Intent, Channel, Technique, Action, Persona)
Indicator types — bank accounts, crypto wallets, emails, phones, etc.
Datasets — restrict results to specific ingestion datasets
Time range — presets for 7d, 30d, 90d windows
Entity filters — match specific entity values with configurable match modes (exact, contains, prefix)
Active campaigns — filter by linked campaign
Click a badge to modify or remove it.
Text search drives semantic retrieval; filters constrain the structured slice. The console merges both streams in the results.
Saving searches
Enter text (optional) and apply any combination of filters.
Click Save search and provide a name.
Manage saved searches in the list below the results — toggle favorites, rename, or delete.
Reading results
Each result card shows a merged score and confidence percentage.
Expand a result to see classification analysis, entity details, and few-shot examples.
Search history is accessible alongside saved searches for quick re-runs.
Last updated