A lot of people have strong opinions on AI and machine learning but getting to speak to real practitioners is a rare opportunity. On Wednesday 20th November, Flexys invited guests from the banking, utility and financial services sectors to join an expert panel for a discussion about artificial intelligence and the role it can play in debt management.
The panel consisted of
Paul Clarke, CTO of Tandem Bank.
Kit Wilson, Head of Transformation at Dŵr Cymru Welsh Water.
Jim Smith, Professor in Interactive AI at UWE (Bristol).
Pedro Ferriera, Machine Learning Research Consultant at Flexys.
Dr David Scholefield, Chief Information Security Officer at Flexys.
It’s all about the data
One thing is for sure, the audience left buzzing with interest and wanting to know more. It’s a huge topic to cover in one evening but we began by discussing everyone’s big concern—data. Do we have enough good quality data? How do we comply with regulations? How can we identify and manage bias?
Our guests were a good reflection of the wider collections community, with some organisations enthusiastically embracing machine learning and some just beginning to consider the journey. They were able to draw on the panel’s first-hand knowledge and experience, with questions flying back and forth all evening.
When it came to practical experience, Paul Clark said that in order to use AI, organisations needed to provide a value exchange, with the benefit to customers being better and more targeted service that is smoother and always up to date. Kit Wilson told the audience that Welsh Water had learned that an expert partner was critical and that any use of AI had to be carefully aligned to business knowledge. Welsh Water, he said, was using machine learning to transform to a pro-active organisation with more insight into its customers.
The panel concluded by offering some thoughts on what guests should take away from the session. They were encouraged to embrace change, to accept that machine learning is an ongoing process and to take account of compliance issues. Paul Clark remarked that the value of machine learning was in the quality of the model and the selection of data features. Finally, Kit Wilson said the AI journey needs to be as safe as possible, bringing security people on board with the exciting changes that had already brought benefits to his organisation and its customers.
Hear more on the forthcoming It’s maths, not magic podcasts
It was great to see so much enthusiasm for the benefits machine learning can offer and to hear how collections professionals are keen to push the boundaries to improve customer service and make more of their resources. We look forward to continuing the conversation and, if you couldn’t make it to the roundtable but would like to hear more, listen to our Refreshing Collections Podcast: It’s Maths, Not Magic: real-world AI for debt management – opportunities, risk and security.