Many banks have started using artificial intelligence to enhance customer service, whether through chatbots, customer acquisition or money management capabilities. However, when it comes to using AI in the backend of banking, those capabilities remains relatively unexplored, except for the most common use cases such as security, AML, KYC or fraud.

One area where AI could come in particularly handy is streamlining employee workload. For instance, using AI to underwrite a corporate loan could free the human underwriter to work on more complex items, such as an SMB loan, allowing the bank not just to expand its products, but also to generate more returns. But there are also simpler, less obvious areas where AI can greatly benefit a bank. French bank Crédit Mutuel is among the few that are currently experimenting with that.

The France-focused bank today announced a partnership with IBM Watson, in which it will be using AI solutions in the backend of all the bank’s businesses: consumer credit, personal protection, payments and risk & compliance.

“This is the first largely deployed AI project undertaken by a large bank that is B2E-centric,” Jean-Philippe Desbiolles, VP of cognitive solutions France at IBM told Bank Innovation. “The point is to use AI to free employees to focus on more precise tasks but also save them the time they spend on unnecessary ones.”

For instance, Desbiolles said Crédit Mutuel “receives a lot of emails, some 350,000 daily.” These emails range from general inquiries to complaints to insurance price quotes. Obviously, tending to these emails is a cumbersome job, so IBM will apply AI in this area and use the technology to categorize these emails, determine the sender’s intent and level of urgency, and then direct them either to a human advisor or an AI-focused assistant, depending on the situation.

This solution, starting today, is available for all employees in the bank’s 5,000 branches. The initiative also includes IBM working with Credit Mutuel’s fintech arm Euro-Information to create and deploy virtual assistants to tackle frequently asked questions and issues across departments like insurance, automotive, savings, health, and pension funds.

“And that’s just one example how AI will be used internally by the bank’s employees,” Desbiolles said.

Indeed, applying AI for internal use can cut costs and equip employees to better serve customers. U.K. challenger Metro Bank Digital Director Alex Park previously told Bank Innovation that Metro Bank is currently exploring ways it can incorporate artificial intelligence for internal use, particularly in handling and understanding data. He also agreed with Desbiolles that AI will help human employees channel their talent into more complex tasks by using AI to streamline certain parts of the job.

Additionally, as a part of this project between IBM and Crédit Mutuel, the two have set up what Debiolles calls a “cognitive factory,” which employs 100 of IBM’s AI engineers and Crédit Mutuel business experts to identify problem areas in which AI can be applied. This is an ongoing project.

In a statement emailed to Bank Innovation, Nicolas Thery, chairman of Crédit Mutuel CM11 Group, said:

Through this new joint-construction stage, Crédit Mutuel is enabled to transform its productivity gains and growth opportunities and pave the way for new services for millions of Crédit Mutuel customers and members.