AMLYZE, the RegTech company trusted by fast-growing fintechs and digital banks, today announced the launch of Sama™, an AI-assisted name-matching capability built into its screening engine. Sama is a purpose-built machine-learning model, developed in-house by AMLYZE’s data science and engineering teams, that recognises genuine name matches that conventional matching logic can miss – helping compliance teams reduce false positives without loosening their controls.
Sama is available now as an opt-in capability across AMLYZE Customer Screening and Payment Screening, and works with an institution’s existing screening configuration.
The problem Sama solves
Name screening forces a difficult trade-off. Loosen the matching and false positives pile up; tighten it and genuine matches slip through – especially when names are transliterated, reordered, merged or misspelled. Analysts then spend most of their time clearing noise instead of investigating real risk.
Sama applies machine learning directly to that trade-off, adding a signal that reads name variation the way an experienced analyst would.
How Sama works
Sama scores how closely an input name matches a watchlist name, at the word level. It works alongside AMLYZE’s existing multi-layered matching engine as an additional signal, strengthening the engine rather than replacing it.
Every match is shown with a detailed breakdown: Sama’s word-level scores, the resulting name-match score, and the engine rules applied on top of it – giving analysts a clear, reviewable view of how the result was reached. The analyst always reviews the alert and makes the final decision.
Sama’s score alongside the conventional engine score. Illustrative example with demo data.
Today Sama is applied selectively, on a controlled and limited set of matches where it is reliable and safe to do so, with its scope widening as confidence grows.
Two ways to use it
Each institution chooses how much weight Sama carries, so teams can adopt AI at their own pace:
- Advisory – Sama’s result is shown as additional information beside the existing match score, helping the analyst decide. Scoring and alert generation are unchanged, and the conventional engine stays authoritative.
- Primary – Sama’s name-match result becomes the effective name score that feeds the final match and alert generation. Even here, the analyst always reviews the alert and makes the final decision.
Teams can start with Sama as a safe, non-intrusive second opinion and move it into the primary role once they trust it.
Key benefits
- Fewer false positives – analysts focus on real risk instead of clearing noise.
- Stronger detection – catches more genuine name variations, including transliterations, that exact matching can miss.
- Detailed, reviewable results – word-level scores and the factors behind each match.
- Adopt at your own pace – from advisory second opinion to primary scoring.
- Part of a multi-layered engine – one more precision mechanism across customer and payment screening: sanctions, PEP, adverse media, counterparty and payment purpose.
Building AI-assisted name matching
We set one condition before putting an AI model into the AMLYZE platform,” said Gabrielius Erikas Bilkštys, CEO and Co-Founder at AMLYZE. “AI in AML only earns trust if a compliance officer can see why a match was scored the way it was and explain it to a regulator. That is the standard we set for Sama, and the standard we will hold for everything we build next.”
Sama is a purpose-built model trained for one job: judging how closely two names refer to the same person or entity,” said Dominykas Stankevičius, Head of Data Science and Analytics at AMLYZE. “It works at the word level, judging how closely two words refer to the same name – transliterations and spelling variants included. That signal then feeds the engine’s existing matching layers, which handle the structure of the name.”
Part of the AMLYZE screening engine
Sama extends the screening module AMLYZE developed in-house and launched in July 2023, now delivered as two products: Customer Screening and Payment Screening. The engine screens against more than 100 authoritative global sources for sanctions, PEPs and adverse media, offers 40+ tunable parameters so institutions can match screening precision to their risk appetite, and maintains a false positive rate below 1.00% in production environments.
Screening is fully connected to the wider AMLYZE platform: payment screening results feed directly into Transaction Monitoring and AML Investigations, enabling automatic holds, enriched alert context and faster resolution from a single integrated compliance platform.
About AMLYZE
AMLYZE is a Software-as-a-Service (SaaS) RegTech company dedicated to merging regulatory precision with modern technological innovation. The company provides a comprehensive suite of tools for financial institutions, including real-time and retrospective transaction monitoring, customer risk assessment, AML/CFT investigations, and PEP, sanctions and adverse media screening.
Trusted by nearly 30 financial institutions across two continents, including Advanzia Bank and EMBank, AMLYZE supports EMIs, digital banks, fintechs and payment businesses in staying ahead of evolving regulatory requirements while enhancing operational resilience. AMLYZE’s products were developed by a team of former regulators and AML experts with extensive experience supervising financial institutions and working with central banks and law enforcement agencies.
AMLYZE has been recognized as one of Europe’s 250 fastest-growing startups by Sifted in 2025, backed by the Financial Times. In the Sifted 250 list, AMLYZE ranked 181st, achieving a 132.89% two-year compound annual growth rate (CAGR) – a testament to its rapid expansion and growing global footprint.
AMLYZE has been working on its core solutions since 2019 and started offering them in autumn 2022. The company has been repeatedly recognized for its innovations in the field of combating financial crime. AMLYZE is the author of AMLTRIX – the community-driven framework designed to standardize the global fight against money laundering by providing a unified approach to detecting and mitigating illicit financial activity.
