Cold Spray Additive Manufacturing (CSAM) is a well-established technology that has recently attracted interest for forming 3D shapes in a fast and scalable fashion. Nonetheless, the resulting surface of cold sprayed parts normally requires post-deposition machining to achieve the desired surface finish. In this work, a convolution-based digital framework for CSAM yield and surface finish prediction able to calculate the optimal interline distance to reduce surface waviness was developed. The aim is to minimise post-deposition treatments, thereby reducing production time, material waste and costs. This method is applicable beyond CSAM and can be of interest for other additive manufacturing techniques.