Adam Gent

A platform for musicians to create and organize setlists around their existing Google Drive files. Solves the problem of scattered charts, recordings, and notes across multiple folders by providing a lightweight wrapper that integrates with Google Drive to keep track of songs, sets, and bands. Musicians can build setlists with everything in one place without moving or changing where their files already live.

A production app that automates concert discovery by scraping your favorite local venue lineups and building auto-updating Spotify playlists. Connect your Spotify account, pick your favorite venues, and get playlists with tracks from upcoming artists. Built with a scalable architecture separating the data pipeline (venue scraping) from playlist generation (SQL + Spotify API) using Python and PostgreSQL.

Predicts traffic collision frequency at intersections using only aerial imagery and infrastructure data, achieving 79% accuracy (R²=0.79). Scraped 1,731 high-resolution aerial images from Google Maps and combined with 20 years of San Francisco collision data. Used transfer learning with ResNet50 for image analysis and created a multi-modal architecture merging visual embeddings with tabular infrastructure features. Key finding: built environment accounts for nearly 80% of collision frequency variation, demonstrating that infrastructure design plays a massive role in traffic safety.

A digital farmhand for wholesale produce management designed specifically for small farms. Features real-time inventory tracking that updates automatically when orders come in, professional mobile-friendly order forms for wholesale customers, automated communications with instant order notifications to farmers and automatic invoices to customers, smart harvest lists that aggregate all orders, and sales analytics with graphs and tables to visualize buying trends. Automates routine tasks so farmers can spend more time in the field.