Built a cloud-based job scheduling system using AWS Services (Kinesis, Dynamo, Fargate) which has been adopted for a range of internal/external uses, including unit testing, documentation building, RNA folding simulations, and user data analysis.

Developed 10+ interactive data analysis apps (Mathematica) including peak-picking for NMR, data clustering for flow cytometry, and image processing for dynamic foam analysis.

Developed unit testing system for interactive Mathematica apps which emulates UI interactions (button presses, sliders, etc.), reducing the need for human-driven QA and saving ~20 person-hours/week.

Developed a novel self-consistent field theory (SCFT) for predicting the structure and properties of polymer electrolytes for Li-ion batteries.

Implemented a high-performance convex optimization package in Fortran 95/C++ (9 months, ~2500 lines) which uses the aforementioned SCFT to efficiently screen candidate materials for desirable properties.

Further improved optimization package by designing spectral algorithm which achieves up to 30x speedup by exploiting symmetry in SCFT equations.

Treasury/Chief Investment Office: Portfolio Strategy

Built stochastic model which generates a distribution of future macroeconomic paths (i.e. time evolution of unemployment, inflation, etc.), conditionally on historical economic data.

Cleaned and compiled macroeconomic data, used to construct regressions of interest rates on macroeconomic variables.