All projects

Projects / Overview

NutriAI - AI-Powered Meal Plan Generator

Built NutriAI: a full-stack, AI-powered meal planning application—in under 12 hours during HackHive 2024. Integrated a local Microsoft Phi-3.5-mini-instruct LLM with a Node.js/Express backend and React frontend to transform complex user dietary requirements into personalized, structured meal plans.

Node.jsExpress.jsReact.jsMicrosoft Phi-3REST APIs
NutriAI - AI-Powered Meal Plan Generator preview

Overview

Developed an agile full-stack web application designed to eliminate dietary planning friction by converting user macro goals and dietary restrictions into actionable meal plans. Built rapidly during a 12-hour hackathon at HackHive 2024, the project emphasizes optimized prompt delivery, clean API orchestration, and component-driven UI design.

Key Features & Engineering

  • Rapid Development & Execution: Designed, coded, and deployed the full-stack architecture in under 12 hours as part of a two-person team at HackHive 2024.
  • On-Device/Targeted LLM Integration: Integrated Microsoft Phi-3.5-mini-instruct to generate structured dietary recommendations based on individual health profiles, caloric targets, and restriction parameters.
  • RESTful Backend Service: Built a lightweight Node.js and Express.js API to validate user inputs, handle asynchronous AI inference requests, and return structured JSON responses to the client.
  • Modular Component Architecture: Developed an accessible React.js interface styled with CSS Modules, prioritizing component reusability and an intuitive, mobile-responsive user workflow.