Although data practitioners have several options for processing structured data at scale, they have fewer tools for processing large corpora of images and videos (e.g., for computer vision use cases). Rikai enables ML engineers and researchers to work with images and video more efficiently (e.g., without writing/managing lots of boilerplate and low-level code). In addition, it includes native support in Spark and PyTorch for images and video to provide a more streamlined workflow from feature engineering in Spark to training in PyTorch.