A team applied a Latent Dirichlet Allocation (LDA) topic model within Natural Language Processing (NLP) to extract and tag maintenance data of 7,000 pieces of equipment. The tagged data were used to train machine learning (ML) algorithms. The initiative was a component of a larger project in the retail environment to develop a predictive maintenance tool using ML. The team accomplished the programming and training in less than 5 days and the LDA model completed the task in 5 minutes on a laptop. This presentation will demonstrate how the technology is maturing and how it can be applied more broadly across multiple areas, functions, and projects.