Project Story: ChefMate
About the Project
ChefMate is a chatbot designed to help users create perfect meals tailored to their ingredients, specifically for one person. The inspiration for this project stems from my personal journey of living alone. I often struggled with cooking the right portion sizes, which led to wasted food and money. I wanted to solve this problem not just for myself but for others facing the same challenge.
The primary goal of this project is to reduce food waste by suggesting recipes based on available ingredients while providing accurate portion sizes for a single serving. ChefMate offers users the recipe details, ingredient quantities, the cuisine type, and step-by-step cooking instructions.
Inspiration
The idea for ChefMate was born out of a very personal need. Living alone, I found it difficult to estimate the right amounts of ingredients for meals, leading to overcooking and discarding leftovers. I realized that many others share this problem, and with rising concerns about food waste globally, I saw an opportunity to create a solution that is both practical and impactful.
How We Built It
Technologies Used:
- RAG (Retrieval-Augmented Generation): Used to enhance the chatbot's ability to generate accurate and diverse recipe suggestions.
- Snowflake: Leveraged for data storage and efficient querying of recipe and ingredient databases.
- Mistral Model: Implemented for natural language understanding and conversational responses.
- RAG (Retrieval-Augmented Generation): Used to enhance the chatbot's ability to generate accurate and diverse recipe suggestions.
Process:
- Gathered a database of recipes with ingredient details, cuisine types, and cooking instructions.
- Designed the chatbot to accept user input of available ingredients and portion size.
- Developed a recommendation engine to match recipes with user-provided inputs.
- Added features to calculate exact ingredient quantities for one serving.
- Gathered a database of recipes with ingredient details, cuisine types, and cooking instructions.
Challenges We Faced
- Ingredient Matching: One of the initial hurdles was designing a system that could accurately match user inputs with recipes, even when ingredients were listed in different formats.
- Portion Calculations: Adapting recipes for single
Log in or sign up for Devpost to join the conversation.