One query, every source
Operators search across connected rosters and registries in a single pass instead of opening five systems and copying values between them.
Fuzzy and phonetic matching
Transliteration variants, alternate spellings, reordered given names and date drift are surfaced as candidates rather than silently missed.
Confidence scoring
Every candidate carries a percentage score with the fields that drove it, so linking decisions are reviewable long after the case closes.
Null results are a valid outcome
When no record exists, operators file a documented null result describing which datasets were swept and why the search terminated. That is a defensible answer for the requesting party - and it stops staff from inventing weak matches to close a task.
Search tiers
Crawl depth and sweep speed scale with the workspace tier assigned to each operator.
Assisted crawling
Higher tiers run parallel sweeps with phonetic expansion; the standard tier uses basic machine crawling.
Every query logged
Terms, datasets touched, duration and result are stamped to the audit trail.
