Projects
Here are some of the projects I've worked on. You can click on the titles to view the source code in GitHub.
Toronto Airbnb Linear Regression Model Price Predictor
Technologies: Python, PyTorch, scikit-learn, NumPy, Matplotlib
This is my work with Torch and scikit to build a linear regression model that predicts Toronto Airbnb nightly prices. The model was trained locally on an M4 Pro MacBook Pro using the Apple MPS backend for PyTorch acceleration. After cleaning the dataset, the script split the data into: Train: 10,866 rows Validation: 2,330 rows Test: 2,329 rows Training stopped automatically at epoch 106 due to early-stopping on validation loss. Final evaluation on the test set gave the following metrics (on the original $ scale): MAE $56.01 RMSE $101.75 R^2 0.4565
TrailSense
Technologies: Typescript, Python, Node.js, Flask, TwelveLabs Pegasus/Marengo
TrailSense is a semantic video search engine for mountain biking trails. Riders can describe the vibe they’re after (e.g. “fast desert trail with berms”), and the system returns relevant clips, difficulty ratings, and location data pulled from crowdsourced trail footage.
SyllaBuddy
Technologies: Typescript, Python, Next.js, Flask
Syllabuddy is a web app that allows users to upload their syllabi for AI to parse and summarize. Key details, dates, and information is extracted, which is then aggregated and displayed to the user. Textbook ISBNs are parsed and will determine external links to download them.
Interested in collaborating? Let's connect!