One of the largest barriers to entry in machine learning is the scarcity of data. Several steps are involved in machine learning model creation, particularly, data collection, data preparation, choosing a model, training, etc. The project focuses on the step of data preparation, namely data augmentation, and specifically in the area of image processing. However, the framework proposed is extensible toward other data type, such as text, and video. Data augmentation is the creation of data from primary data samples using well-established transformations, such as rotations, pepper-salt distortions, etc. to create more samples for training a machine learning model. More notably, serverless computing, which involves a platform distributing processes across different transient servers, is increasingly a popular method in computing as it streamlines the setup of acquiring and establishing hardware, networks, etc. The project proposes framework that establishes a programming interface for serverless functions to be called to process the data.