![](https://i0.wp.com/finovate.com/wp-content/uploads/2023/10/pexels-cottonbro-studio-6195484.jpg?w=640&ssl=1)
- Featurespace unveiled its generative AI-powered, Large Transaction Model (LTM), TallierLTM, this week.
- The technology uncovers hidden transactional patterns typically undetected by current methods that may be indicative of criminal activity.
- Featurespace made its Finovate debut in 2016, appearing at both FinovateEurope and FinovateFall that year.
Fraud and financial crime prevention company Featurespace unveiled its Large Transaction Model (LTM), TallierLTM, this week. A foundational technology for payments in specific and the financial services industry in general, TallierLTM is a large-scale, self-supervised, and pre-trained model built to power the next generation of AI apps to protect consumers from financial crime.
The technology marks the first time financial professionals in the fraud fighting space have been able to leverage a generative Large Transaction Model. Featurespace noted in a statement that TallierLTM has provided improvements of as much as 71% in fraud value detection compared to the industry standard.
“What OpenAI’s LLMs have done for language, TallierLTM will do for payments,” Featurespace founder David Excell said. “There is widespread concern about how deep-fakes and generative AI have been used to deceive consumers and our financial systems. We plan to reverse this trend by utilizing the power of generative AI algorithms to create solutions that protect consumers and make the world a safer place to transact.”
Connecting to FIs via its enbedding API, TallierLTM analyzes billions of transactions, identifying hidden transactional patterns that current methods often cannot detect. The technology’s insights are based on time sequencing, discovering unusual spending patterns over a short period of time, for example, or between a consumer and a merchant. This increased ability to distinguish legitimate activity from potentially criminal behavior will not only enable data scientists to improve their model’s performance faster, Featurespace Chief Innovation Officer Dr. David Sutton said. It will also allow institutions to “realize the value of machine learning investments more quickly.”
“We know that smarter technology helps financial institutions better understand their consumers,” Sutton added. “We have taken this to the next level by pairing cutting-edge generative AI algorithms with huge volumes of data, enabling a machine to efficiently comprehend the relationships between different customer transactions.”
Founded in 2016 and headquartered in Cambridge, U.K., Featurespace made its Finovate debut in 2016, appearing at both FinovateEurope and FinovateFall. An innovator in adaptive behavioral analytics and automated deep behavioral networks for risk management, Featurespace serves more than 80 direct customers and 200,000 institutions. In recent months, the company announced partnerships with digital payment platform Clip and fellow Finovate alum Zeta. In August, Featurespace launched its ARIC Scam Detect solution to help protect financial services companies and their customers from scams – especially Authorized Push Payment (APP) scams – in real-time.
“As scammers become increasingly sophisticated in their tactics, with the use of Generative AI and machine learning, FIs need an adaptive solution that can protect from changing scam types in real time and monitor both inbound and outbound payments,” company Chief Product Officer Pat Hinchin said.
Featurespace has raised nearly $108 million in funding from investors including Chrysalis Investments, MissionOG, and Insight Partners.