A man faces a computer screen with algorithms and data used for research. — Dr. Akash Deep, an assistant professor in the School of Industrial Engineering and Management, utilized machine learning in a project to optimize business processes with Jewelers Mutual.

IEM research uses machine learning to aid business process decisions

Wednesday, July 2, 2025

Media Contact: Tanner Holubar | Communications Specialist | 405-744-2065 | tanner.holubar@okstate.edu

Research using machine learning and data analytics spans a variety of industries, from infrastructure to health care. Researchers in the College of Engineering, Architecture and Technology at Oklahoma State University’s School of Industrial Engineering and Management look to employ machine learning and data science to improve the business process for Jewelers Mutual ^®.

This project is part of a collaboration with the University of Wisconsin-Madison and OSU, working with Jewelers Mutual. OSU’s primary contribution is creating a custom business credit score model to be added to the company’s tool belt. 

The collaboration has spanned multiple research projects, with a major emphasis on loss prevention for the company’s customers. Through studying historical data on crime-related losses, safety measures and store locations, researchers are identifying patterns and risk factors to improve the company’s underwriting process and customer guidance.

Dr. Akash Deep, an assistant professor in IEM, has worked through this collaboration for five years, with this being the latest undertaking. He was co-advised for his Ph.D. at Wisconsin by Dr. Raj Veeramani and Dr. Shiyu Zhou. Deep began working on this collaboration before becoming a faculty member at OSU in 2022.

A man in a portrait photo wearing a nice dark suit. — Dr. Akash Deep.

“The overall goal for this project is to see how we can use machine learning models and data science insights to improve business processes,” Deep said.

The custom business credit score model will improve risk assessment decisions in the company’s underwriting process. They will create the model using third-party data, such as credit-based consumer information from major credit bureaus.

There are three primary goals for this research: 1) Define improvement opportunities and then discover how data science can alleviate it. 2) Develop and apply data science and machine learning to address the opportunity, which includes analyzing the data, building models, and creating a prototype. 3) Share the findings with Jewelers Mutual for them to potentially apply to the company’s business practices.

“Typically, we can claim to be the experts in data science, but of course the company is the expert on their business,” Deep said. “We start with descriptive analytics to try and understand what we can through the data. We want to be treating their data properly to optimize their business effectively.” 

Deep’s team would like to reach the point in the research where certain outcomes such as losses can be predicted. It is a collaboration that Deep has enjoyed being a part of in multiple research endeavors. 

“I’m pretty lucky to be able to collaborate with Jewelers Mutual,” Deep said. “It has been an interesting experience personally. You know, I can do a lot of math, create models, but the real value comes when it is used for a specific and beneficial reason. 

“It is a fantastic project to work on to really see how we can help. It is a nice, synergistic collaboration — working with business experts and very smart students, I think it has been a very great experience altogether.”