Research and Software
← All projects
Extracting Cognitive SLUMS Scores from VA Clinical Notes
Rule-based NLP pipeline to extract Saint Louis University Mental Status (SLUMS) scores from ~3 million unstructured VA clinical notes. 83% accuracy, 99% precision, 20,000+ notes/minute on a single laptop. Presented at GSA 2025.
2025
PhD Thesis: NLP for Uncovering the Illicit Massage Industry
PhD thesis (Harvard, Sep 2023) applying NLP and finetuned BERT models to classify online massage parlor reviews, enabling geographic analysis of illicit activity to support counter-trafficking research.
2023
Sampling-Based Path Planning Algorithms
Implementing and evaluating RRT and PRM motion planning algorithms from scratch in Julia, with analysis of how clutter affects success rates. Rotation project reviewing FMT*.
2020
Ocean Buoy Path Planning Simulation
Using public ocean current data to build a simulation framework for testing motion planning algorithms on autonomous ocean buoys. Solo contract work in Python, Pandas, and Folium.
2020
Digger Finger: GelSight Tactile Sensor
A miniature wedge-shaped GelSight visual-tactile sensor specialized for identifying objects buried in granular media. Published at ISER 2020. Featured on MIT News.
2020
ICRA Paper: Six-Axis Force-Torque Sensor
My paper on designing an inexpensive six-axis force-torque sensor was accepted to ICRA 2020. 🎉
2020
Robot Compiler (RoCo) Website Redesign
Redesign of the UCLA LEMUR Lab RoCo webapp to use Bootstrap for mobile/desktop responsiveness, plus bash/Python/OpenSCAD scripting to generate STL thumbnails for the carousel.
2019
CS279r: PaperSignals Parameterized
An open-source browser-based UI (Python/Flask) for customizing Google's voice-controlled printable paper robots. Harvard PL/HCI Graduate Seminar.
2018
CS223: Credit Community Graph Algorithms
Investigating trust-score propagation in social lending networks to establish creditworthiness without traditional data. Harvard Random Processes and Algorithms.
2018