Finds identities hidden in large volumes of text and data.
Matches them against lists using 11 identity variant models across 15 languages.
Uses machine learning rather than name matching lists for better adaptability.
Designed to integrate into AML solutions, on premise or in the cloud.
Used by banks and border security agencies around the world.
8 Questions to Ask About Identity Matching
- How reliable is your identity matching function?
Rosette identifies people through their legal name, nicknames, initials, honorifics, misspellings, swapping of name components, and date of birth, regardless of language.
- Can your solution match identities even when name components are entered into the wrong fields?
Sometimes name components are transposed. Mexican surnames, for example, are often two words. A common mistake is to enter one of them as a middle name. Rosette accounts for these conditions because it is culturally aware.
- Does your solution match identities in Russian, Arabic, Japanese, Korean, and Chinese?
Rosette provides cross-lingual matching and linguistic expertise in difficult, non-Latin languages.
- Can your solution identify Abu Mohammad Al Julani, Mohamed ElJulani, أبو محمد الجولاني, 阿布·穆罕默德·約拉尼, and Абу Мухаммад аль-Джулани as the same person?
Most technologies can’t because they resolve matches based on lists, and if the match is not in the list, it’s not found. Rosette uses machine learning-based identity matching so it can match identities encountered for the first time.
- Can your solution track identities from all sources typically used in business?
Rosette can extract identities from virtually any source: documents, webpages, wire transfers, teller systems, brokerage and trust databases, invoices, contracts, Internet news articles, and social media.
- Can your solution accommodate watchlists from other countries in other languages?
Rosette can because it can match identities across languages.
- Can your solution scale affordably as identity matching demands continue to grow?
All watchlists are works in progress, so they grow. This is a problem for list-based identity matching because the number of name variations to match in the list grows exponentially with each new name. Since Rosette uses machine learning-based matching, it can absorb the growth with minimal hardware and cost.
- Is there a way to increase accuracy and cross-lingual matching functionality in the solution you have now?
Yes. Rosette is designed to seamlessly integrate with your existing functionality without adding a lot of additional hardware.