Chipotle Cultures
Team consisting of UCLA ME & Data‑Science William Tjandra (Python, SQL, XGBoost, Node.js) and NYU Langone researcher Hubert Leo (full‑stack JS, Angular, Ionic, C++).
Project Description
Chipotle Cultures is a Chrome extension that reimagines Chipotle’s menu with authentic cuisines (Italian, Chinese, Jamaican, Bronx, Brooklyn, English). A dropdown switches cuisines and the entire UI—hero sections, items, images, and descriptions—updates dynamically. Each item has an info button that fetches cultural context and prep methods via the Tavily API. A voice-interactive ordering system (with a talking avatar) guides users through protein → rice → beans → toppings with cuisine-specific prompts. An Express.js backend uses Redis caching and connection pooling to persist food-info lookups, cutting Tavily calls by 90%+ for fast, low-cost responses.
Prototype: Fully functional—dynamic cuisine switching, custom imagery, Tavily-powered info, and voice-guided ordering with working avatar.
Tech: Chrome Extension API, Node.js, Express.js, Redis (caching + pooling), Tavily API, ElevenLabs API (optional voice), HTML5 Audio API, IntersectionObserver, MutationObserver.
Prior Work
Prior Work:
- Chrome extension skeleton with basic image replacement.
Hackathon Work:
- Multi-cuisine system (Italian, Chinese, Jamaican, Bronx, Brooklyn, English) with dropdown switching.
- Dynamic UI swap: hero sections, item names, images, and descriptions per cuisine.
- Info buttons → modal with cultural context & prep methods via Tavily API.
- Express.js backend with health checks + Redis connection pooling.
- Redis caching layer for food-info queries (reduces Tavily calls by 90%+).
- Voice-guided ordering (protein → rice → beans → toppings) with cuisine-specific prompts (ElevenLabs optional).
- Organized assets by cuisine; complete image pipeline and config built during the hackathon.