Open Meal Buddy: A Local AI Meal Planner for Someone You Love

I developed a local AI meal-planning tool called Open Meal Buddy for my roommate. It takes into account their allergies, pantry items, food preferences, and weekly budget to generate practical meal plans without sending any personal data to a third-party server. My goal was to make healthy, affordable cooking simpler for someone who is busy, has dietary restrictions, and desires an uncomplicated meal plan.

You can access a working local demo at: http://127.0.0.1:5000. To run the tool from the terminal, use the command: python app.py --name Ava --allergies peanuts,dairy --budget 40 --days 3 --pantry rice, eggs, spinach, chickpeas, onion --likes vegetarian, quick meals --schedule busy weekday dinners. The AI model used in this project is Ollama, an open-weight local model.

The Python app builds a personalized meal plan from real-life constraints by structuring the prompt with the user's allergies and food restrictions, pantry ingredients, weekly budget, preferences, and schedule constraints. This design allows the system to be easy to run, customize, and private by default. Open innovation plays a significant role in this project as it ensures the experience remains private, affordable, and under the user's control.

By running the app locally on a laptop, the user's dietary history and food preferences stay within their own device, eliminating the need to send sensitive data to a remote server. This approach also reduces API costs for each generated plan, makes it easier to swap and fine-tune the model later, and ensures the tool remains functional even without an internet connection for day-to-day use.

My project was built with a practical focus on a real person and a real problem: to simplify healthy meal planning while addressing allergies, budget, and time constraints. I competed in the Hacktoberfest Weekend Challenge, aiming for two prize categories: Best use of open-source AI and Best project built for a friend or loved one.