Data science is a field that focuses on developing methods of recording, storing, and analyzing data to effectively extract useful information. It continues to be one of the largest growing, most sought after fields in the technology industry, with positions in analytics growing 150% in 2019. The field combines components of computer science, statistics, and business expertise to take on some of the industry’s most challenging problems, including medical imaging, genomics, preventative medicine, and diagnoses.
I began my path to becoming a data scientist in college. I originally entered as a Biochemistry major but found the mathematical portions of Biology far more interesting and added a second major in Applied Mathematics and Statistics. I then moved on to complete my master’s degree in Applied Mathematics where I took courses in Data Science, Machine Learning, and Computational Modeling. I loved these courses because instead of doing standard textbook problems, I solved real world problems. Using mathematics and advanced statistical models to build products and offer insights into complex problems thrilled me.
Softheon’s Data Scientists are part of a larger Data Governance team, composed of Data Scientists, Database Administrators, Software Engineers and Technical Product Managers. Our team tracks Softheon’s data lifecycle, working to ensure the accuracy and integrity of our data. We then use the data to make key insights into our company, members and the healthcare field.
One of the projects I most recently worked on was a Plan Recommendation Engine. We noticed that sometimes new members may have difficulty finding the right health plan for them, so we wanted to make the process a little easier.
Our Plan Recommendation Engine takes in information, such as a member’s age and zip code, and uses it to predict, with over 99% accuracy, the top three plans he or she may choose. This information is then presented to the member and helps narrow down the plans to save each member time and energy when making their decision. The model uses a Random Forest algorithm, which is a machine learning (an application of artificial intelligence that provides systems the ability to automatically learn and improve from experience without being explicitly programmed) classifier, combining thousands of decision trees into a forest.
Along with machine learning models, our team also works on dashboards to visually track and analyze metrics. I recently worked on a dashboard that looks at what time of day members make their insurance payments, with respect to their time zone, as well as the payment history of members who were either cancelled or terminated from their insurance coverage. We used this information to see if there was a trend in members making payments during certain hours of the night and them being termed from coverage later in the year. If we can find indicators of unhappy members or members who may be forgetting their payments early on, we can provide these insights to the issuer for them to do outreach and retain membership.
Some of our future projects include building out a recommendation system for our enterprise web platform and incorporating both image and voice recognition into our enrollment processes.
Data science and predictive analytics are valuable tools that are revolutionizing the healthcare industry. The more information we receive from our issuers, the more accurate models we can make, providing better insights into the community we serve. These insights are incredibly valuable in helping us create better products and achieve higher key performance indicators. With the exponential growth in the quantity of data being created, companies who aren’t investing in data science are going to be left behind.