🌟 Inspiration Small bakeries often face the challenge of balancing production and demand. Producing too much leads to waste, while producing too little results in missed opportunities. We wanted to build a smart solution that helps bakeries make data-driven decisions about daily bread production — that's how Nom Pang Maeo was born.
⚙️ What it does Nom Pang Maeo is a sales forecasting tool that predicts the number of breads to bake each day. By analyzing past sales data, seasonal patterns, and other relevant factors, it helps bakeries optimize inventory, reduce waste, and improve efficiency.
🛠️ How we built it We used the following technologies and steps:
Data Collection: Cleaned and preprocessed historical bread sales data.
Modeling: Implemented machine learning models such as XGBoost and Prophet for time series forecasting.
Evaluation: Compared models using MAE, RMSE, and R² to choose the best performer.
Visualization: Created graphs to show actual vs. predicted sales trends.
🚧 Challenges we ran into Handling missing or inconsistent data.
Choosing the right model that balances accuracy and simplicity.
Dealing with unpredictable factors like holidays or promotions affecting sales.
🏆 Accomplishments that we're proud of Successfully reduced prediction error to a practical level for real-world use.
Built a flexible system that can be adapted to other bakery products.
Learned to balance technical complexity with usability for non-technical users.
📚 What we learned Practical experience in applying machine learning to real-world business problems.
How to handle and clean time-series sales data.
The importance of understanding user needs when building AI-powered tools.
🚀 What's next for Nom Pang Maeo Integrate with POS systems for real-time prediction updates.
Add features like weather or holiday impact to improve forecast accuracy.
Build a user-friendly web dashboard for bakery owners to track and adjust their production plans easily.
Log in or sign up for Devpost to join the conversation.