Seneca News

Debt collection research involves looking at a range of data. (Adobe Stock)

Applied Research project aims to improve debt recovery

Seneca teams up with a fintech to help predict the propensity to repay loans

April 29, 2024

It’s a problem that has plagued lenders for a long time – how do you collect overdue debts?

Seneca’s Applied Research and s2h, a fintech, software innovation company, are tackling the issue by developing a program using artificial intelligence to help predict the propensity to repay and improve debt collection.

"This is at the leading edge of debt collection," said Dr. Mariam Daoud, Professor, School of Software Design & Data Science and Principal Investigator for the project.

The research team is analyzing data, writing code and training various models to predict debt recovery. They are working with the company’s D2R-Collect product, which enables cloud-based technology to automate administrative tasks, so lenders can collect money faster and at higher rates of recovery.

Sam Andary, the CEO of s2h, which is based in Markham about 30 kilometres northeast of Toronto, said it’s a new era for technology in debt collection.

"Recovery of debt has become a high priority. You can imagine the urgency to get paid for products or services in today’s economy.

"The research so far has been extremely helpful,” Mr. Andary said, noting successful collection requires more than credit rating scores.

Sam Andary, CEO, s2h

The researchers have been looking at everything from debtor’s responses to emails and phone calls to the level of debt and the amount of time various loans spend in collection. They analyze the information, work with a number of algorithms to get the best data modelling and assess results for accuracy, Dr. Daoud said.

The research is funded by the Natural Sciences and Engineering Research Council of Canada. And the initial project, from September 2022 to June 2023, went so well that follow-up research with more specific demographic data began in January and will run through June 2024. In total, six research assistants from the Honours Bachelor of Data Science & Analytics (DSA) program are involved.

The assistants present the research to the D2R-Collect team weekly and Mr. Andary said he’s been impressed with their professionalism and progress.

Dr. Mariam Daoud
Dr. Mariam Daoud, Professor, School of Software Design & Data Science

Dr. Daoud agreed. 

“Having Seneca students to work on the project is one of the best things,” she said. “It adds experiential learning to their career and gives them the opportunity to apply their coding and critical thinking within a real project and contribute to delivering business solutions.”

Aliya Kerimbekova, a fourth-year DSA student who is working on the project, said it’s a great experience. This is the second research project she’s worked on at Seneca during her studies.

“Doing research projects helps you lock in the knowledge you learn in school,” Dr. Kerimbekova said.

It also gives you an opportunity to explore different things and learn more, she noted.

Currently, she’s doing data analysis and creating scatter plots and histograms as part of the research to see the relationships between variables. She’s particularly interested in predictive models and she’d ultimately like to work in machine learning engineering.

Aliya Kerimbekova, Research Assistant and Data Science & Analytics student

Dr. Daoud said the type of work they’re doing can easily be applied to other sectors such as sales or health care.

For now, the goal is to help get D2R-Collect ready for commercialization.

“It has the potential to significantly improve debt collection,” Dr. Daoud said.