Head of Data Science at Unum

Reed Hayes is an accomplished leader in the field of data science, currently serving as the Vice President and Head of Data Science at Unum. With expertise in deep tech, machine learning, and AI, Reed has worked with several organizations to develop innovative products, including founding Rendever (Virtual Reality for Seniors), a cutting-edge itinerary planning engine (Disney Genie), and AI-based insurance claim adjudication and retention engines. Reed's key strengths lie in his idea instincts and technical creativity to bring together diverse ideas and translate them into strategic plans that can be effectively implemented for organizational transformation. He has a track record of solving very ambitious and technically challenging problems at companies like Coca-Cola, Disney, Rendever, and Unum.

How would you describe your journey with Unum, and what are your key responsibilities here?

Unum is a Fortune 500 insurance company that provides comprehensive insurance products to a broad spectrum of clients across various industries. We constantly strive to leverage cutting-edge AI/ML techniques to generate value across the insurance value chain.

As the VP and head of data science, my primary role is to create ambitious AI/ML based projects, develop key AI/ML strategies, and lead a large technical team to deliver on our AI/ML vision. In addition, I supervise various projects and initiatives, coordinate teams, and manage stakeholders and executives to ensure we are working on the right things.

Please shed some light on the significant challenges and trends existing in the insurance landscape today.

 The increased use of technology has improved productivity and efficiency in all aspects of the insurance industry 

The insurance industry is typically considered more conservative compared to sectors such as automotive or entertainment. However, in the past five years, there has been significant adoption of machine learning technology to address growing market needs. Machine learning technology is now being used in various aspects of insurance businesses, including underwriting processes, risk calculation, claims adjudication, sales improvement, and customer retention. The increased use of technology has improved productivity and efficiency in all aspects of the insurance industry.

Are there any recent projects you have been working on, and what are some of the process elements leveraged to make them successful?

We have recently been working on a customer retention project aimed at improving customer loyalty. Many players in the insurance industry are attempting to build churn or probability models that can predict customer loss. However, unlike others, we have developed a holistic solution by combining a churn model with reinforcement learning. This allows us to determine the optimal intervention for preventing loss and retaining customers. Our solution can analyze a customer's service, billing, or pricing issues and identify the root cause of the problem.

In addition, we have built an engine that can predict when customers are at risk and, more importantly, suggest the most optimal course of action to save them. This project has been highly successful in preventing customer loss, which has resulted in increased customer satisfaction and retention.

In light of your experience, what will be your advice to fellow peers in the industry?

Every aspiring data science professional has to remain curious and ambitious toward solving problems that come their way. When I initially design an AI/ML solution, I attempt to create the design that will not just be the perfect answer 1 year from now but 15 years into the future. In addition, I highly recommend data scientists to learn how to blend multiple different quantitative fields together to create the perfect answer (Economics, Machine-learning, Optimization, Game-Theory, etc).