Industry work at Block and TruStage Insurance. Details are described at the level I can share.
Headcount Forecasting Pipeline
Block · 2025
Built an end-to-end pipeline to forecast incoming headcount for Block's Finance and Strategy department. Achieved greater than 85% accuracy, enabling the team to plan resourcing ahead of quarterly cycles without relying on vendor tooling.
Python
Databricks
Forecasting
Employee Survey LLM Analysis
TruStage · 2023
Fine-tuned a large language model to analyze tens of thousands of open-ended employee survey comments. Reduced the time from data collection to actionable insights by approximately 80%, making this kind of qualitative analysis feasible at organizational scale.
LLM Fine-tuning
NLP
Python
Sales Performance Driver Model
Block · 2025
Built a predictive ML pipeline for Block's global Sales organization to identify the key factors driving sales and revenue performance. Output was used to inform compensation and management strategy across multiple regions.
Python
Sklearn
Databricks
Engineering Health Metrics Pipeline
Block/Square · 2024
Designed an end-to-end data pipeline for the Square Engineering function linking system outages to operational response and stabilization timelines.
Salary Modeling Application
TruStage / Block
Developed a custom compensation modeling application integrating payroll data and linear regression outputs to support executive salary decisions. Reduced dependence on external vendor tools, saving approximately $60K in annual licensing costs.
Python
Regression
Databricks
Student Selection Dashboard
Winfred Arthur · 2019
Redesigned the veterinary school's student admissions dashboard (VBA and Excel) for a university client.