Built book recommendation system using data extracted from 26m Amazon book reviews and 6m Goodreads reviews. Used collaborative filtering via implicit ALS algorithm to build ensemble of 5 models on user:user, user:item, and item:item vectors. Ongoing NLP work on reviews for topic modeling. (AWS, Big Data, MongoDB, PySpark, PyTorch, Spotlight, Implicit, Cython, OpenMP, NLTK)