AMSTERDAM — The algorithms are starting to add up at Temenos.

The Geneva, Switzerland-based banking technology company will add up to 15 algorithms for artificial intelligence before yearend. This goes with the 20 or so AI algorithms the company already has available for its clients. Not all of those 20 algorithms, however, are available for live use – “in production,” in industry parlance. But they represent a notable library of algorithms, nonetheless.

The number of algorithms matter because each represents a capability that can be used to assess data. In short, the more data that flows into AI algorithms, the better the machine-learned results of those algorithms. With each algorithm, banking progresses toward an increasingly automated paradigm.

In an interview with Bank Innovation here today, Prema Varadhan, a chief architect at Temenos and leader of the team working on developing AI algorithms for the company, said Temenos has identified 15 additional use cases for AI that are part of its “big pipeline” for this year. Each of those use cases could turn into an AI algorithm, she said.

“The more models we build, the more accuracy we get,” she said.

The AI efforts are being undertaken by Temenos’s AI Centre for Excellence, which develops use cases for applying AI “at scale.”

Of the 20 AI algorithms Temenos has in hand today, the two most popular among its clients are related to customer attrition and funds flow. The customer attrition algorithm allows banks to use machine learning to understand or forecast which clients are leaving the bank. The funds-flow algorithm, meanwhile, can be used to predict how funds will flow into an account in the future. The algorithm can predict when a banking client — consumer or business enterprise — might be overdrawn or will miss direct payments. Varadhan said banks can build proactive push products using such algorithms.

Varadhan said that Temenos often uses open source AI code to build its banking-specific algorithms. While she did not specify which open-source offerings the company uses, some popular AI code providers include TensorFlow, Keras and Scikit-learn.

Temenos (OTC: TMNSF) has a market capitalization of about $10.9 billion.