<p>This study explores the use of Convolutional Neural Networks (CNN) to automate milling process planning, a task that is traditionally complex and time-consuming. Process planning, which converts engineering drawings into manufacturing operations, involves steps such as feature recognition, operation sequencing, tool selection, tool path planning, and setting cutting parameters. Existing methods like optimization algorithms and expert systems face limitations in handling the full complexity of this task. In addition, most research focuses on only one step rather than providing an integrated system to fully automate the process. The purpose of this study is to develop a CNN-based system that can perform complete process planning from CAD files with minimal human intervention. This research shows that CNNs can effectively automate these processes with high accuracy. The approach uses cross-sectional images of Computer-Aided Design (CAD) files, similar to 3D printing layer slicing, to capture part geometry and generate process plans. By integrating advanced model architectures, including Residual blocks from ResNet and Inception blocks from GoogleNet as feature extraction layers (FELs), the developed CNN model achieved an accuracy of 94.72% in inferring operation sequences, tool selection, and tool path patterns for 2.5D parts at once. The system can infer planning results in just 3&#xa0;s and generate the complete tool path and numerical control code within minutes when used in conjunction with Computer-Aided Manufacturing (CAM) software. This significant reduction in process planning time minimizes the workload for engineers. The results demonstrate that CNN-based systems have the potential to revolutionize process planning by improving efficiency, reducing lead times, and decreasing the need for human intervention in routine planning tasks.</p>