Fridger.ai

Fridger.ai turns a snap of your fridge into instant, nutrient-rich recipes. Save hundreds, slash up to 40% of food waste, and shrink your carbon footprint, one meal at a time.

Comment

Inspiration

We never knew what good foods to make at home, and ended up eating out a lot, wasting food and money.

What it does

Using Gemini AI, generates recipes based on uploaded photos from your fridge, using no extra ingredients unless stated otherwise.

How we built it

Using Python, Next.js, Tailwind, MongoDB, Pydantic, Gemini Vision, Spoonacular API, built an AI Agent using Google Gemini to process ingredients from a fridge, then generates recipes from those ingredients using Spoonacular API. Then displayed those recipes in a dashboard that could be saved or deleted in a user's profile.

Challenges we ran into

Had trouble with authentication, storing user data & recipes, and AI prompts for accurate ingredient identification.

Accomplishments that we're proud of

We are proud of building a strong AI Agent and ensuring precise image and recipe analysis and generation.

What we learned

We learned how Agentic AI can be used to solve problems, gained a deeper understanding of image processing, database control, and git team workflow.

What's next for Fridger.ai

We plan to implement a mobile application to make image uploading smoother and more available for clients. We also plan to make the AI model more efficient at crafting recipes and adding more ways for users to input ingredients that do not require refrigeration. We also plan to consider dietary restrictions.

Built With

Share this project:

Updates