projects
Leftovers is a web application that generates recipes based
on user-inputted ingredients, leveraging LLMs to deliver dynamic results.
Built in TypeScript/React.
The KCL Diagnostic Tool is a tool that allows users of the
Kinesis Client Library to diagnose and troubleshoot issues with their KCL applications.
Developed during my AWS Summer 2025 Internship.
Built in Python with CloudWatch/DynamoDB.
The Heart Failure Analysis project is a comprehensive data analysis and visualization
initiative focused on heart failure patient data from the UCI Machine Learning Repository. The project
includes the full development pipeline, from data cleaning and preprocessing, exploratory data analysis (EDA),
and dashboard generation with sample prediction.
Built in Python with Pandas/Plotly/Dash.
Calculadora is a browser-based calculator application, capable of basic mathematical
operations. Inspired by the iOS calculator app and developed as part of The Odin Project curriculum.
Built in HTML/CSS/JavaScript.
Etch-A-Sketch is a browser-based drawing application, allowing users to
the cursor to draw on an increasingly opaque canvas. Inspired by the classic toy
and developed as part of The Odin Project curriculum.
Built in HTML/CSS/JavaScript.
The Spam vs. Ham classifier is a logistic regression model,
capable of discerning between spam emails and legitimate emails. The accompanying web application
takes a user-inputted email to predict its legitimacy.
Built in Python with Flask.
The Natural Disaster Classifier is a set of computer vision models, trained to classify and
differentiate between natural disaster images and their apparent damage level. Developed as the final project
for UC Berkeley's Data 200S: Principles and Techniques of Data Science.
JujuType is a speed typing game inspired by Monkeytype and my favorite anime, Jujutsu Kaisen.
Built in HTML/CSS/JavaScript.
Simple Substructure Search is a Python module that processes molecular
structural data files (.sdf), returns their graphical representations via NetworkX, and can
detect the presence of substructures via molecular fingerprinting.
