Coffee beans are one of the most important agricultural commodities in the world. An accurate process of classifying coffee beans is key in ensuring that only high-quality beans reach consumers. Classifying coffee beans is not simple because the process relies on human judgment, which is prone to errors and inconsistencies. Errors in classification will result in a decrease in the value of coffee bean products. In this context, the role of information technology in automating and improving the accuracy of the classification process is urgently needed. This research method is the use of one of the digital technologies, namely CNN for classifying coffee beans, with the aim of determining the effectiveness of using CNN technology to classify coffee beans and knowing the effectiveness of implementing project-based laboratory learning. The results obtained show that the use of CNN has effectiveness with an accuracy rate above 90%, showing its advantages compared to other neural network methods. Then, the effectiveness of using information media that utilizes CNN technology on student learning outcomes was recorded to be very satisfactory, with effectiveness reaching 93.60%.