Many governments, corporations, and non-profit organizations have a strong motivation to protect rainforests and take action against deforestation, particularly if this deforestation occurs as a result of exportable commercial commodities such as coffee or cocoa. However, there have traditionally been significant limitations in measuring and detecting coffee region deforestation at scale. We have developed a more accurate approach to detecting deforestation that addresses these limitations due to recent advances in satellite imagery and machine learning, and welcome collaboration with coffee companies. Such a rigorous assessment of deforestation in the coffee sector has not been carried out before. We believe that this innovation has the potential to become an important new tool for the coffee sector in its efforts to combat deforestation and mitigate climate change. It can not only underpin future research but also has important policy implications for organizations on the ground. Indeed, in the context of an expanding regulatory environment led by a more demanding civil society, it provides an accurate, consistent, and transparent way for organizations to report on deforestation events in their supply chains and monitor them. At the same time, we hope that it will open a broader discussion regarding the potential for machine learning to apply new innovations to systemic problems that plague the coffee sector.