Inspiration

As college students, we're always juggling the need for affordable, convenient meals with the desire to eat healthily. Too often, the quest for convenience leads to overspending or compromising on nutrition. Recognizing this common dilemma, we sought to create a solution that balances cost, convenience, and health.

What Meal Planner Does

Enter Meal Planner: a bespoke meal planning service tailored for budget-conscious, nutrition-savvy individuals. By simply entering your dietary preferences, budget, and any additional notes, Meal Planner crafts three personalized meal plans. These plans are not only nutritious but also designed for easy purchase at your nearest Wakefern superstore, ensuring both convenience and affordability.

How We Built It

Leveraging the Wakefern API, we accessed a comprehensive database of store items, which we categorized by food type for easy processing. Our system employs OpenAI's GPT-3.5 Turbo to interpret user inputs—including dietary restrictions, budget, and personal notes—and match them with suitable food items. The platform, deployed via Streamlit, offers a seamless, user-friendly interface.

Challenges We Overcame

A significant hurdle was ensuring consistency in the AI-generated meal plans, as output formats were initially varied. By fine-tuning the model's parameters and setting a strict output format, we significantly improved the user experience, ensuring clear, concise meal plans every time.

Our Proud Accomplishments

We are thrilled with Meal Planner's ability to deliver consistent, precise pricing within its meal plans, including thoughtful recommendations for common household items like salt and pepper. This not only demonstrates the model's nuanced understanding of culinary needs but also enhances practicality for users.

Lessons Learned

This project was a valuable lesson in applying AI models to specific domains, showcasing the potential of artificial intelligence to revolutionize everyday tasks like meal planning.

What's Next for Meal Planner

Looking ahead, we aim to refine our model's accuracy by incorporating domain-specific embeddings. Furthermore, we plan to integrate Meal Planner directly with Wakefern's shopping system, allowing users to automatically generate shopping carts based on their meal plans. This next step will bridge the gap between planning and purchasing, making Meal Planner an indispensable tool for healthy, budget-friendly eating.

Built With

Share this project:

Updates