Utilized python to build a data pipeline to prepare data and extract relevant features from a popular music website. Applied python with relevant APIs to collect 4GB+ worth of audio data for extraction of several spectral features into raw dimensions to form a dataset. Implemented an unsupervised machine learning technique to extract repeated audio patterns from multiple song tracks and reduce high dimensional data from audio features, which reduced the original dimensionto half.Delivered finished designs to clients via Built and evaluated multiple models using several machine learning algorithms with team members including implementing an Artificial neural network. Deployed model to a flask web app and hosted on a cloud platform.