Portfolio

For my convenience as I apply for jobs, I developed my own resume generator. It requires a JSON containing the resume information, and renders a nicely formatted PDF resume in the format I previously worked with in Google Docs.

I worked alonside two classmates and good friends, Pramit Vyas and Tyler Quinn, to deliver a data solution to the United States Tennis Association (USTA) to help them identify and predict the performance of promising American junior tennis talent. As a massive fan of tennis for 10 years at the time of writing this, I was obviously extremely excited to work on this project. My expertise in professional tennis certainly helped us tremendously as we worked on this project to understand the direction of the project as well as the data and other resources at our disposal. ...

The goal of the project was to investigate whether the time of a course affected student performance. I designing a blocking experiment to isolate the effect of course time on performance by designing the experiment to optimally control for instructor, student year, race, gender, and prior GPA. This project consisted of the following steps: Designing the experiment Writing the proposal Performing the analysis Writing the report

I completed this project alongside for classmates from my graduate program cohort. We sought to predict electricity prices from historical electricity price data, AI investment data, and Google Trend results for AI. Ultimately, we found that a SARIMA model was sufficient for accurately predicting electricity prices, and the AI data was not particularly useful for predictions, although certain limitations in our analysis exist. Together, we delivered a presentation and submitted a report showcasing our results.

This final project, which I completed alongside my classmate and good friend Pramit, consisted of two parts: Predicting flight delays to and from Pittsburgh using past flight data Writing up results in a report Besides the flight data we were given, I gathered and tried to utilize other data, namely list of major American Federal holidays, historical weather rain and wind data, but none of these methods improved predictive accuracy by much. I expect that the weather data could have been more useful if I had historical forecasts, not just observed weather, but I could not find any. ...

Data analysis report investigating causal claims about the effectiveness of various types of HIV treatment on post-treatment outcomes. Read here

I completed this project alongside four classmates from my graduate program cohort. We built an end-to-end data pipeline and interactive dashboard to process and visualize large-scale public higher education datasets from the U.S. Department of Education’s College Scorecard and IPEDS. We designed a custom PostgreSQL schema and ETL upsert pipeline to handle multi-year institutional metrics, financial aid data, and enrollment figures. Ultimately, we integrated this database into a dashboard that updates given the data (must be run locally). ...

Infographic depicting the unprecendented domination of the Big 3/4 in Men’s professional tennis through various key statistics and records.

Data analysis report investigating whether proximity to rail trails affects home value. Read here

Optimization project for managing CMU Orientation food logistics. View PDF

Reddit linguistic analysis across generations. I particularly loved setting up the data pipeline. View PDF

I created a website for the CMU Pickleball club which I founded. To be quite honest, it was mostly vibe coded in pure HTML, CSS, and JS. It is hosted on GitHub Pages. Website

AP Research paper on using various gamification and types of leaderboards to enhance learning/performance in a novel cognitive task. View PDF