Meal Tracker

Inspiration

As college students, we struggled to stay consistent with nutrition while managing heavy homework and deadlines. Meal planning felt like one more decision-heavy task at the end of long days, so we built Meal Tracker to automate it and keep us aligned with our goals.

What it does

Meal Tracker helps users set calorie and macro targets, log meals, scan food with AI for estimated nutrition, and generate a full weekly meal plan. It also provides recipes, grocery lists with estimated costs, shopping directions, and weekly planner integration.

How we built it

We built the app as a lightweight web platform using HTML, CSS, and vanilla JavaScript on the frontend, with a Node.js + Express backend for API routes. We integrated OpenRouter for AI-powered chat, food image analysis, and meal planning, USDA FoodData Central for nutrition search, Supabase for authentication/data sync, and deployed on Vercel.

Challenges we ran into

Our biggest challenges were reliability and quality control around AI outputs. We had to handle malformed/incomplete JSON, enforce structured responses, and improve retries/fallbacks so the app remained stable. We also had to tune prompt quality to get practical results (like realistic shopping locations and useful plan details), while balancing development time with school workloads.

Accomplishments that we're proud of

We’re proud that Meal Tracker solves a real problem we personally faced. We shipped an: goal setup, logging, AI scan, weekly planning, grocery support, and planner import—all in one app.

What we learned

We learned that building with AI is as much about system design as it is about model prompts. Strong validation, parsing, and UX feedback are essential. We also learned how to scope features, iterate quickly, and prioritize user value under tight time constraints.

What's next for Meal Tracker

Next, we want to improve personalization and accuracy: smarter meal-plan adaptation over time, better grocery deduplication and store matching, stronger nutrition estimation from scans, and richer progress analytics. We also plan to refine collaboration features and continue improving reliability for daily use.

Built With

• c
• cmake
• css
• ctest
• express.js
• html
• javascript
• node.js
• npm
• openrouter
• supabase
• supabase-progres
• usda-nutrition-data
• vanilla-js
• vercel