Cognitive business strategist Michel Van der Poorten explains IBM’s approach to the development of cognitive business systems and the huge impact it will have on both business models and customer experience. 

 

What is cognitive banking?

It is important to clarify what is meant by cognitive. At IBM, this means augmenting human intelligence rather than replacing it – we develop systems that help people take better decisions. Our system encompasses four new capabilities: understanding human language and tools; reasoning and drawing hypotheses; learning though human validation or rejection of such hypotheses; and interaction with humans because this system is not a black box. The overall approach is to process complex structured data with unstructured data, such as legal text, graphs or images, and provide informed answers. Our system will help legal experts prepare for upcoming regulations by understanding the operational intricacies and facilitating the linkage between regulation and compliance. Client behaviour will be easier to predict, because our cognitive system will be able to cross-check past actions and provide predictive models. Risk assessment is another application, making credit scoring for corporate loans easier and enabling banks constantly to reassess risk profiles.

 

« Banks using cognitive technology will gain a competitive advantage, and open a gap to those that don’t.” »

 

 

How will cognitive banking change the business landscape?

Employees will strongly be impacted by the cognitive system. By freeing up time spent on data analysis, experts will be able to focus on their core work, and the tool will augment their expertise. In addition, the client experience will be improved because knowledge will be delivered to them more quickly and of much higher quality, enabling organisations to be proactive rather than reacting responsively to consumer needs and expectations. Cognitive systems will also have considerable business impact in that banks using the technology will gain a competitive advantage, and open a gap to those that don’t. Complex data processes will become efficient and benefit from an improvement in quality.

 

What are the development challenges linked to your system?

Our projects start with a design thinking workshop to determine the best use case with which to start the cognitive build-up. Typically, we define the customer experience the organisation wants to achieve and identify the data sources – internal, but also external content such as Reuters newsfeeds – as well as assess the quality of the data. Surprisingly, there are no real IT challenges. The complexity lies in merging data such as logs from a call centre with first-hand expert knowledge that we need to train the cognitive system. Currently experts are feeding the system as an auxiliary task to their main work. As the sector realises the importance of cognitive processing, future job profiles will emerge with the sole purpose of training, testing and evolving these systems.