๐ฝ๏ธ MyMealAgent
๐ Inspiration
MyMealAgent was born from a simple but universal problem:
We save dozens of recipes from cookbooks, YouTube, TikTok, and social media - but most of them never make it to the kitchen.
The gap isnโt inspiration.
Itโs execution.
Planning meals, checking ingredients, writing grocery lists, and managing time often break the flow between โThat looks deliciousโ and โDinner is ready.โ
MyMealAgent exists to close that gap.
๐ง What It Does
MyMealAgent transforms saved recipes into real meals.
It intelligently considers:
- ๐ฅ Ingredients users already have
- โฑ๏ธ Available cooking time
- ๐จโ๐ฉโ๐งโ๐ฆ Number of servings
- ๐
When the meal is planned
Then it:
- ๐ณ Selects the best recipe option
- ๐ Generates a smart grocery list
- โฐ Schedules cooking reminders
- ๐ฉโ๐ณ Guides users step-by-step through the cooking process
From inspiration to execution - seamlessly.
๐ ๏ธ How We Built It
๐ฑ Frontend
- React Native (Expo)
- Clean modular architecture
- Production-ready iOS build via TestFlight
- RevenueCat for subscriptions
โ๏ธ Backend
- FastAPI REST APIs
- Celery for asynchronous recipe imports
- Redis for task queue
- Docker for containerized deployment
- Azure for cloud hosting
๐ค AI & Processing
- Tesseract OCR for extracting text from images
- OpenAI API for:
- Converting video to transcript
- Summarizing transcript into structured ingredients & steps
- Intelligent parsing of social media recipe URLs
๐งฉ Challenges We Ran Into
- ๐ฑ Building, deploying, and testing our first iOS app with Expo + TestFlight
- ๐ฅ Handling large video URLs and bot detection from YouTube & Instagram
- ๐ผ๏ธ Processing large images reliably with OCR inside Docker containers
- โ๏ธ Deploying Celery workers and Redis in cloud infrastructure
- ๐ณ Integrating RevenueCat with App Store subscriptions
๐ Accomplishments Weโre Proud Of
- Successfully processing large images and long-form video URLs
- Building a fully async recipe import pipeline
- Deploying scalable Docker infrastructure with OCR support
- Getting in-app purchases working via RevenueCat Test Store
๐ What We Learned
- End-to-end iOS development with Expo
- FastAPI backend architecture with async task processing
- AI-driven text extraction and summarization workflows
- Docker-based deployment for cloud-ready services
- Subscription handling with RevenueCat + TestFlight
๐ฎ Whatโs Next for MyMealAgent
- ๐ Support for more website and recipe sources
- ๐ฝ๏ธ Enhanced AI recipe optimization
- ๐๏ธ Improved grocery intelligence
- ๐ณ Production-ready App Store subscription rollout
- ๐ค Social and shared meal planning features
๐ก The Vision
MyMealAgent is not just a recipe organizer.
Itโs an intelligent cooking companion that helps people turn saved inspiration into real meals - effortlessly.
Log in or sign up for Devpost to join the conversation.