Sheetz Cooler Optimization

Overview

As part of a team project, I worked on a confidential computer vision–driven solution addressing operational challenges in cooler restocking for Sheetz. The problem centered around limited real-time visibility into cooler stock levels, leading to inefficiencies in restocking workflows and missed sales opportunities. Our team designed and prototyped a system leveraging computer vision techniques to monitor cooler inventory states and surface actionable insights for restocking, while respecting operational constraints and confidentiality requirements. The solution focused on improving accuracy, reducing manual checks, and enabling faster decision-making for store staff. My contributions spanned solution design, technical exploration of CV approaches, and translating operational pain points into a scalable system concept. This project emphasized applying machine learning to real-world retail environments, balancing technical feasibility with business impact.