Automated business processes, modeled mortality trends during COVID-19, and trained large language models to predict cause of death codes.- Analyzed excess death during the COVID-19 era and found 10,000+ more deaths than expected from historical trends, mostly in low-vaccination areas.- Delivered conference presentations and a written report to disseminate excess death findings- Trained a large language model for death ICD-10 coding that achieved 90% accuracy in test data.- Developed procedures to securely geocode addresses of 11.8 records.- Utilized address correction API to automate resolution for 29% of data entry errors.- Automated the delivery of 90+ reports for federal, local, and private partners.

Provided database design and coding support to an experimental risk stratified payment model for mental health services. Evaluated novel behavioral health programs.- Developed dashboards and other visualizations to track progress on key performance indicators such as timeliness of data and hospitalization trends.- Cleaned, analyzed, and mined data in support of a new payment model.- Wrote clear technical documentation of new procedures and reviewed these documents with a mental health provider network.- Predicted which clients would use crisis services 5+ times in the next six months using random forest models.

Streamlined and expanded analysis tools for state programs to provide payment assistance for drugs to prevent and treat HIV/AIDS. Improved reporting pipelines.- Analyzed charge codes, CPT codes, prescription utilization, and diagnosis codes in a state HIV medical claims program.- Quantified staffing time required to address processing backlogs in medical claims.- Analyzed claims data to identify why dental claims costs were growing quickly.- Performed probabilistic matching of client data using Link Plus, SAS, and R.- Mapped county-level incidence of sexually transmitted infections using ArcGIS.

Improve analytical pipelines via automation. Performed literature reviews, scientific writing, and peer reviews to advance behavioral health science in traffic safety.- Improved efficiency of data pipelines with SQL-like SAS procedures.- Wrote methods section of scientific article on traffic safety in rural and urban areas- Performed logistic and Poisson regression analyses- Used multiple imputation to handle missing survey response data.- Performed qualitative analysis of short answer responses to evaluate an educational pamphlet