Ripjar, the provider of smarter screening and risk intelligence solutions, has introduced new AI innovations within ULTRA, the real intelligence engine that powers screening operations at financial institutions and enterprises, including 25% of the Global Systemically Important Banks and 35+ of the Global Fortune 500. The updates enable compliance teams to identify and manage risk with greater confidence, speed and clarity, making screening more intelligent, explainable and actionable at scale.

Built on national security grade technology, ULTRA delivers a level of accuracy and efficiency that off-the-shelf tools and generic LLMs simply cannot match. Powered by proprietary graph algorithms, transformer models and custom AI, the engine resolves more than 10 billion historic articles and over 5 million new articles a day, across more than 150 languages, into a single system of record for entity risk.

By blending multiple data sources, including PEPs, sanctions, watchlists and adverse media, into one resolved profile, users gain a complete 3D view of risk. Risk lands on the right entity, with 98.7% accuracy, reducing the volume of alerts reaching an analyst by 99%. The engine continues to learn, so an analyst resolves an entity once and gets the benefit for the lifetime of that customer relationship.

The latest innovations include enhancements to Screening Assistant, Ripjar’s agentic AI capability for alert triage, to further reduce routine alert handling, improve decision-making and increase risk coverage. Updates to ULTRA’s entity matching means customers can now better distinguish between parent organisations, subsidiaries and similarly named group entities. This removes guesswork and strengthens Screening Assistant’s ability to mimic human reasoning when deciding whether or not an alert relates to the exact entity. The update also extends Screening Assistant’s handling of complex naming conventions, including Arabic names, to support more global use cases.

Ripjar has also enhanced ULTRA’s adverse media risk classifiers, which analyse each news article to determine if it relates to a specific risk and identify the person involved. The latest model upgrades ULTRA to a new generation of machine learning, improving the quality of existing classification and entity-resolution tasks.

ULTRA is proven in highly regulated and mission-critical environments where explainability and governance are vital. Every alert, decision and summary is backed by evidence, traceable to its source and held in an auditable system of record built for regulatory scrutiny.

Conrad Nicholas, Chief Product Officer at Ripjar, said: “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. It works like a funnel, with each layer building on the last. What sets ULTRA apart is orchestrating these models at scale with explainability built in, so customers can trust the outcomes as well as benefit from more efficient screening.”

Matt Mills, CEO at Ripjar, said: “Knowing who you’re doing business with is becoming more difficult. Sanctions change rapidly, adverse media multiplies and AI is enabling criminals to become more sophisticated. AI is vital but compliance teams need to understand and trust its decision making. We understand the importance of this in part due to our background in the UK’s intelligence services. ULTRA is the result of years of engineering and proven deployments and these latest innovations take explainability and performance to new levels.”

About Ripjar:

Ripjar provides customer screening, adverse media monitoring and threat investigation software to 400+ organisations globally, including Tier 1 banks, enterprises and government agencies. Customers include 25% of the Global Systemically Important Banks and more than 35 of the Global Fortune 500.

Founded by former UK intelligence service technologists, Ripjar is built on national security grade technology to help organisations transform AML compliance and combat financial crime. Its screening solutions are powered by ULTRA, Ripjar’s real intelligence engine, proven in highly regulated environments where explainability and governance are vital. ULTRA blends billions of data sources including PEPs, sanctions and adverse media articles into one resolved profile, so risk lands on the right entity.

For more information, visit ripjar.com and follow @Ripjar on LinkedIn.

Media gallery

About The Author
Business Wire