Nathan White Email & Phone Number

Greenwich, Ct, Us

Teaneck, New Jersey, Us

 Build predictive models using Python to predict medication adherence for Health First members. Used Tableau to provide medication metrics to internal clients at Health First.  Developed machine learning models with Aetna Active Health Management to determine the likelihood a member will complete the Health Risk Assessment. Utilized Google Cloud Platform used to conduct the analyses. Develop predictive clinical algorithms for the SAG AFTRA Health Plan to help reduce healthcare costs and predict eligibility.

Hartford, Ct, Us

Managed a team of data scientists responsible for building and maintaining predictive models for group benefits insurance. Collaborate with stakeholders in GB Actuarial and Marketing to determine the predictive models needed solve pricing and operational issues. Designed feasibility studies to determine model priority. Led the predictive modeling effort to build a Quote Viability predictive model to better identify quotes The Hartford is unlikely to win. We can then decline low viability quotes and better direct resources to high viability customers. Managed the effort to get manual rate predictive models for Long Term Disability and Life products into a Model Factory that automatically produces the model scores and gives us diagnostics on how the model is performing. Utilized our existing Model Health Check system to determine when models need to be retrained refreshed. Continuously worked with offshore resources to update the Health Check system. Utilized Python for data mining and predictive modeling work and mentored actuarial students on predictive modeling best practices

Bloomfield, Ct, Us

Manage a team of 4 data scientists responsible for building predictive models focused on customer and clinical outcomes.  Collaborate with stakeholders on the Consumer Health Engagement Team to determine the predictive models needed to identify members for clinical outreach. Manage and maintain all predictive and identification models that contribute to the Health Matters Score to predict the estimated cost savings available to each member. Health Matters is then used to identify members for chronic condition coaching and Case Management. Work on Co-development of an action engine to track the week to week changes in all models that feed Health Matters Score. Maintain the One Guide scoring logic so each member with this app would know the best cost savings opportunities available based on their health needs. Utilize of new predictive modeling efforts such as machine learning in Data Robot and Python and help direct reports use these new methods. Build new predictive models to identify the likelihood an existing health care member will purchase voluntary products such as hospitalization coverage. Take the lead on a new predictive modeling effort with Consumer Health Engagement to predict the mortality rate of members with any one of 424 high risk diagnoses in the past year.

Bloomfield, Ct, Us

Built a new predictive generalized linear model to determine members at high risk for a Short Term Disability. This data was used to get members in wellness programs to prevent the disability claim and reduce duration if a claim occurs. Provide several marketing proof points to show that Cigna disability health care connect business yields a lower disability incidence rate than clients who have just short term disability coverage without Cigna Medical coverage. Development of new statistical models to better target new clients for Long Term Disability and Pharmacy coverage. The models target clients who are likely to buy coverage and clients that will be a low risk. Responsible for implementing the new Health Matters score pilot that combines models developed by Predictive Analytics in a cost savings opportunity. Build a new pharmacy analytics model to determine which members are most likely to switch from a brand name drug to a generic drug when the savings opportunity exists. Mentor new employees in Predictive Analytics for both statistical modeling techniques and accurate data mining specific at Cigna. Provide mapping and spatial data analysis using ArcGIS to better target regions and neighborhoods for wellness program outreach. Build and implement new Volume to Value statistical models that reward health care providers for better health outcomes instead of fee for service. Utilized Oracle, SQL, SAS, SAS EG, and Teradata to extract data and build predictive models.

Windsor Locks, Ct, Us

Analyzed quality data from the Electronic Engine Controls of the PW535, PW617, and V2500. Utilized statistical analysis tools in DERBY, Minitab, and Excel for this data. Developed new statistical models to determine root cause of seal leakage and fuel flow problems on the Jet fuel Control for the PW617.

New York, Ny, Us

Developed new multivariate predictive models to price various small business insurance products including property, liability, auto, and workers comp. Use of SAS and SAS Enterprise Guide to analyze very large data sets. Made use of Tweedie and Frequency – Severity methodologies along with loss ratio and the pure premium approach to develop the models.

- Responsibale for tracking all in game statistics for the UConn Men's and Women's Ice Hockey teams using ice hockey in game computer programs.- Assisted in the in game statistics tracking for the UConn Women's Volleyball team.

East Hartford, Ct, Us

- Assisted in the development of the Pratt & Whitney probabilistic design system and responsible for maintaining the statistics portion of the system.- Developed many new statistical regression models to predict the life of engine parts and to create new powder metal alloys using SAS and Minitab.- Taught Basic Statistics, Minitab, and Six Sigma courses through the Pratt & Whitney and United Technologies education centers.

Intern Statistician

Worked with the Hartford Whalers team statistician on in game statistics, media relations and reporting, and ensuring NHL and Whalers statistics are accurate and in agreement. Worked directly with Frank Polnaszek.

Bachelor Of Science (B.S.), Statistics

Master Of Science (M.S.), Statistics