nlu

Why we included this project

Most NLP libraries force you to tie together pipelines and manage model objects before you can do anything. This one flips that: you call nlu.load() with a model name and apply it straight to a dataframe column, so text processing becomes a matter of a single line instead of boilerplate. It wraps Spark NLP under the hood, giving pandas users access to a large catalog of pretrained components for tasks like entity recognition, sentiment, translation, embeddings, and parsing across many languages. The Streamlit integration is the same idea taken further, letting you spin up an interactive demo of those models with one command. If your team already lives inside Spark NLP, it might feel redundant, but for pandas-centric work that wants broad NLP coverage without the plumbing, it's a genuinely comfortable starting point.