Team Reference Code: 2B
Inspiration
Many students struggle to find cheap alternatives when it comes to food and often resort to buying meals that will cumulatively exceed their budgets. Cooking is the go to cheaper resource when it comes to meals, but many students struggle when it comes to cooking due to the lack of knowledge, time and skills. Hence we invented FrAI_(/fraɪ/)_, your personal AI recipe generator to help you create recipes based on your customisations.
What FrAI can do
FrAI is a personalised AI recipe generator and cooking assistant, designed to assist students in creating budget-friendly and easy-to-cook meals. The platform takes into account individual preferences, providing customised recipes that align with the user's available ingredients and skill level. Users can input their preferences, such as cuisine and ingredients on hand, and FrAI generates tailored recipes, making cooking more simplified for students.
How we built it
FrAI was meticulously crafted leveraging advanced technologies, with a primary focus on the capabilities of the GPT 3.5 model. The backbone of our system is powered by state-of-the-art natural language processing (NLP) and text generation, allowing users to interact effortlessly with the recipe generation process. FrAI hosts three main features: recipe generation, image generation, and a chatbot agent.
1. Recipe Generation:
Utilising GPT 3.5 - Turbo, we implemented a robust text generation system that understands and processes user preferences, allowing for dynamic and personalised recipe recommendations. The model's natural language understanding ensures that user inputs are interpreted contextually, providing more accurate and tailored responses based on user preferences (i.e. complexity, duration, available ingredients).
2. Image Generation:
To enhance the user experience, we incorporated image generation capabilities into our platform. Stable Diffusion assists in creating visually appealing representations of the recommended recipes, giving users a preview of the dish before they decide to proceed. This functionality not only captivates users but also assists those who rely on visual cues when choosing recipes.
3. FrAI Agent Chatbot:
Our chatbot functionality is driven by GPT 3.5, transforming the interaction into a dynamic and informative conversation. Users can engage in a dialogue with FrAI, seeking clarifications on cooking techniques, ingredient substitutions, and recipe adjustments. FrAI understands natural language queries, providing a user-friendly and interactive experience for those looking for cooking guidance.
Challenges we ran into
- Integrating AI algorithms to generate accurate and diverse recipe recommendations.
- Designing an intuitive user interface that caters to a diverse user base.
- Ensuring data security and privacy, especially when dealing with user preferences.
Accomplishments that we're proud of
- Successful integration of AI for personalised recipe suggestions.
- Creating an accessible platform that addresses the cooking challenges faced by students.
- Overcoming design challenges to make the interface user-friendly.
What we learned
- Enhanced knowledge of AI/ML integration in real-world applications.
- Improved skills in frontend and backend development.
- Gained insights into user experience (UX) design principles.
What's next for FrAI
- Implementing user feedback to enhance and refine the platform.
- Integrating additional features such as meal planning and grocery list generation.
- Exploring partnerships with food suppliers or grocery delivery services.
- Building on the success FrAI in the culinary world, we hope to develop Aicohol - an AI-assisted mixologist and bartender, the next evolution of FrAI.
Built With
- axios
- css
- javascript
- node.js
- openai
- react


Log in or sign up for Devpost to join the conversation.