Sales Forecasting Program Utilizing Generative AI for Decision Support in a Local Retail Bakery

Model Development Based on Practical Production Judgment Data

Abstract

This study, in collaboration with a local bakery, developed a machine learning-based sales forecasting program. Traditional forecasts relied on craftsmanship experience, making judgment difficult in the artisan’s absence. By incorporating weather, day of the week, and past production data, the program enables store staff to perform sales forecasting without relying solely on artisan expertise. A graduating student with little coding experience used enerative AI in development, highlighting the potential for participation by non-experts. By attempting to formalize periencedependent decision-making in local retail bakery, the research is situated within the broader context of food loss reduction and may provide foundational knowledge for future initiatives.