ING’s approach to AI explained in eight minutes!

Bahadir Yilmaz, ING’s chief analytics officer sat down with Fintech Finance to detail ways that ING is using AI, machine learning, and advanced analytics.

Not surprisingly, Bahadir said one of the main use cases for AI is the use of chatbots to improve customer service.

“If you look at the whole journey, the goal of the customer is not to call a contact centre. They have a problem they want fixed and chatbots can play a crucial role there.”

“We’ve been playing with chatbots since 2017. The chatbox experience that you had five years ago wasn’t great. It was rudimentary. We noticed this too and quickly shifted our technology from a more statistical natural language processing (NLP)-based approach to a large language models-based approach and we see big improvements.

“Our customer satisfaction levels are all going up. Clients want an instant answer to their question and want an instant solution to their problems. Chatbots are highly capable of that,” he said.

Bahadir said ING also sees potential in using AI and machine learning technologies in know your customer processes, client services, sustainability transition plans, pricing, marketing campaigns and fraud.

Faster and more personalised loans

“We see more and more institutions using the power of transactional data and coming up with better lending propositions. They’re able to provide loans in a matter of minutes instead of days. At ING we are also doing it. Loans are getting faster, more instant, more personalised which is great for customer satisfaction,” said Bahadir.

“You’re also making the whole landing system more inclusive because you take the human element out of it. You’re making it more data driven.”

While the implementation of AI can have huge benefits, Bahadir acknowledged there is anxiety over the use of AI and he himself is also concerned about the limitless application of AI.

“We are planning to alleviate the concerns of customers by implementing robust risk management mechanisms around AI. Introducing generative AI techniques to a business problem is only five percent of the job.”

“Ninety-five percent of the job starts after that. It is important to build systems around AI tools and that takes a lot of effort. At ING, we have a 20-step process which evaluates any AI system for 140 risks,” said Bahadir.