Ripjar boosts AI in customer screening tools
Ripjar has deployed fresh artificial intelligence capabilities inside ULTRA, the real intelligence engine that supports screening operations for financial institutions and enterprises. These upgrades allow compliance staff to detect and control risk with increased assurance, velocity, and lucidity.
ULTRA achieves a precision and productivity level that standard commercial tools and generic large language models cannot provide, relying on national security grade technology. It processes over 10 billion historical articles and 5 million fresh articles daily, covering 150 languages, into a unified system of record for entity risk.
Upgraded Screening Capabilities
This engine combines numerous data streams, such as Politically Exposed Persons, sanctions lists, watchlists, and adverse media, into one consolidated profile, giving users a full 3D view of hazards. Risk is correctly attributed to the specific entity with 98.7% accuracy, cutting the number of alerts sent to an analyst by 99%.
The newest developments include improvements to Screening Assistant, Ripjar’s agentic AI tool for alert triage, which aims to decrease routine alert management and refine decision-making. Changes to ULTRA’s entity matching enable clients to better differentiate between parent companies, subsidiaries, and similarly named group entities.
Read Also: Peak3 Launches AI-DLC Cutting Insurance Tech Costs 50%
Precision Entity Matching and Adverse Media Risk Classifiers
ULTRA’s entity matching eliminates guesswork and strengthens Screening Assistant’s capacity to simulate human logic when determining if an alert concerns the precise entity. The update also expands Screening Assistant’s ability to handle detailed naming conventions, including Arabic names, to support a wider range of global use cases.
Ripjar has also refined ULTRA’s adverse media risk classifiers, which examine each news article to decide if it pertains to a specific risk and identify the individual involved. The latest model advances ULTRA to a new generation of machine learning, boosting the quality of current classification and entity-resolution activities.
According to Conrad Nicholas, Ripjar’s Chief Product Officer, “ULTRA is one engine, but it runs multiple layers of AI, each performing a specific task: stripping out duplicate noise, resolving mentions across billions of articles into single entity profiles, and attributing risk to the right person.”
ULTRA functions in highly regulated and mission-critical settings where explainability and governance are essential. Every alert, decision, and summary is supported by evidence, traceable to its origin, and stored in an auditable system of record designed for regulatory review.
