Trained in-house with tens of thousands of documents classified and processed in 6 different languages, this machine learning model can categorize documents with a 93% accuracy.