Projects

Things I've built and worked on.

BikeShare

May 2026 – Present

Built a full-stack peer-to-peer rental marketplace with JWT auth and role-based access across 3 user roles, 12 pages, and ~30 REST endpoints, letting students list, search, and rent bikes, scooters, and skateboards.
Implemented indexed geospatial radius search in PostGIS over a GiST spatial index, with radius, price, type, rating, and availability filters.
Engineered a privacy layer that exposes only approximate pickup locations until a booking is accepted.
Engineered a booking engine that prevents double-booking under concurrency, combining interval-overlap checks with pessimistic row-level locking to serialize simultaneous acceptances, enforced by a state machine and integration tests.

June 2026 – July 2026

Built an end-to-end CLI that transforms unstructured Shopify catalog data into queryable product facets using deterministic extraction, embeddings, reranking, LLM classification, and computer vision with confidence-based abstention.
Designed a resumable, cost-aware inference pipeline with closed vocabularies, Zod schema validation, granular caching, concurrency controls, exponential backoff, provenance tracking, and offline provider fallbacks.
Implemented automated taxonomy discovery, text-vision fusion, human correction workflows, Shopify metafield exports, and HTML/CSV coverage reports, validated by a deterministic test suite.

July 2025 – September 2025

Built a Purdue dining hall macro tracker that lets users log meals, set macro goals, and view daily calorie and macro totals.
Automated daily ingestion of Purdue dining court menus via Vercel Cron so menu data stays current without manual updates.
Hardened production with Sentry error tracking and structured logging to shorten debugging of failed ingestion runs.

March 2025 – May 2025

Built an LSTM-based poker agent using PyTorch and self-play reinforcement learning, achieving a stable positive win rate against baseline opponents.