A local server designed to load one or more pre-trained NLP models during startup and expose them through a clean, RESTful API