XConnect adds KYC fuzzy matching to cut identity mismatches

XConnect launched a KYC fuzzy-matching feature this week to help identity-verification systems link customer records that differ because of typos, transliteration, nicknames, diacritics or inconsistent formatting.

The capability applies approximate string-matching to name, address and date-of-birth fields. It is delivered through XConnect’s existing API and can be tuned by clients to balance match sensitivity against the volume of manual reviews.

XConnect said the fuzzy matching runs in real time during onboarding and ongoing monitoring flows. The company described the method as a combination of phonetic matching, edit-distance calculations and token-based comparisons that produces a probabilistic match score instead of a strict exact match result.

Firms can set score thresholds so high-confidence matches proceed automatically while lower scores are flagged for manual review. XConnect also noted the feature can assist cross-border checks where names and addresses are spelled differently across alphabets or national conventions.

The vendor presented the tool as applicable to new-account onboarding, transaction screening and identity resolution across fragmented datasets. XConnect highlighted reduced friction for users entering personal data on mobile forms and lower volumes of manual interventions caused by common data-entry errors.

Regulatory requirements for customer identity verification under anti-money-laundering and counter-terror financing rules were cited as background to the launch. XConnect said the feature can be customised by market and regulatory requirement so matching rules align with local compliance practices.

The fuzzy-matching capability is available now to existing customers via the API and will be offered to new customers as part of XConnect’s identity-checking suite. The company did not disclose client names or performance metrics tied to the release.

In a statement, XConnect said, “Fuzzy matching lets our clients accept legitimate customers who would otherwise be delayed by strict exact matching, while still allowing teams to review uncertain cases quickly.”