Experience
Data Science Intern
Metropolitan Transportation Authority (MTA)
New York, NY · June 2026 – Aug 2026
- Modernized legacy Oracle procedures with Spark SQL and Python, integrating them into Apache Airflow ETL pipelines processing ~80,000 operational records per day.
- Tested graph-tool for Apache Spark on Kubernetes as a replacement for NetworkX, projecting an 85% reduction in annual compute costs for the transit ridership model.
Research Assistant — Parallel Computing and ML in Biophysics
Texas Christian University, Physics and Astronomy
Fort Worth, TX · Apr 2025 – Dec 2025
- Used CUDA to accelerate agent-based model (ABM) simulations of viral spread on NVIDIA GPUs.
- Implemented graph neural networks to study spatial heterogeneity of syncytial cells via topological features.
Visiting Internship — Data Analysis for Atomic Physics
Raman Research Institute, Light and Matter Physics
Bangalore, India · Jun 2024 – Jul 2024
- Investigated ultra-cold trapping of sodium and potassium atoms using optical dipole traps.
- Used Python to extract laser beam image matrices and assess Gaussian-fitted spatial profiles, uncovering a millimeter-scale misalignment.
Research Assistant — Materials Science Characterization
Texas Christian University, Physics and Astronomy
Fort Worth, TX · Sep 2023 – Apr 2025
- Built components for an ultrahigh vacuum (UHV) chamber enabling cathodoluminescence spectroscopy of nanomaterials.
- Led a study of preparation techniques for nanocrystalline zinc oxide and gallium oxide.
Featured
Data Science at the MTA
As a Data Science Intern at the Metropolitan Transportation Authority, I worked with the Data & Analytics team on modernizing legacy Oracle procedures into Spark SQL and Python, integrating them into Apache Airflow ETL pipelines that process roughly 80,000 operational records a day.
I also validated large operational datasets with SQL and Trino, tracking down discrepancies to improve the reliability of data used downstream. On the modeling side, I tested graph-tool on Apache Spark and Kubernetes as a replacement for NetworkX in the MTA's transit ridership model, identifying an approach projected to cut annual compute costs for the model by 85%.
I closed out the internship by benchmarking the ridership model's performance across multiple runs and presenting efficiency recommendations directly to Data & Analytics stakeholders.
At the MTA
