SecurePoint Trade
Know whether the hit is real before you stop the shipment.
SecurePoint Trade combines four matching methods into one explainable score so analysts can separate transliteration noise from a real restricted-party risk.
Catch transliterations, typos, reordered segments, and sound-alike names.
Show the scoring logic instead of hiding it behind a black-box match.
Give analysts a visible score breakdown they can use during review.
Entity Resolution Engine
Input entity
MOHAMMAD AL-RASHEED
Phonetic
Muhammed El-Rashid
Levenshtein
Mohamad Al-Rasheed
Jaro-Winkler
Mohammed Rasheed
Trigram
Al Rasheed Mohammad
Composite confidence
92%
See the score inside the live transaction review
This screenshot comes from the real Trade application using sanitized demo data. The analyst can inspect the composite confidence and the underlying match detail before deciding.
- Composite confidence is visible on the record the analyst is reviewing.
- Match context sits beside the decision controls.
- The scoring explanation supports a defensible release decision.
Each layer catches a different kind of name problem
This page should feel technical because buyers evaluating entity resolution want to see the mechanics, not just the promise.
Phonetic Matching
Sound-alike detectionCatches names that sound identical but are transliterated differently across languages and scripts.
Levenshtein Distance
Typo detectionMeasures insertions, deletions, and substitutions to catch data-entry errors and common typos.
Jaro-Winkler
Transposition detectionOptimized for short strings and prefixes. Catches character transpositions and partial matches.
Trigram Similarity
Segment reorderingBreaks names into 3-character segments and compares overlap to catch reordered name components.
All four algorithms converge into
One Confidence Score
A single 0-100% match confidence score tells analysts how likely a hit is to be real, reducing false positives and accelerating clear decisions.
Why It Matters
Entity resolution is where screening becomes usable.
If every near-match lands in the queue, analysts burn time proving noise is noise. If matching is too strict, real exposure slips through. The point is not a fancy algorithm. The point is a defensible decision threshold that your team can actually operate.
4
Scoring layers
0-100%
Unified output
Explainable
Analyst review
FAQ
Entity resolution questions
Focused on matching quality, score interpretation, and analyst trust.
See the score breakdown on a live transaction
The fastest way to evaluate this module is to watch one transaction move from hit to decision with the full scoring detail visible.