We are starting in the world of NLP/NLU in our company. Just a couple of months ago we developed a simple chatbot with Google’s Dialogflow with voice interactions (Speech to text, text to speech). And just a few weeks we are creating training data for our legal department with Spacy and Prodigy. Early-stage.
How is working with RASA in Spanish since we are located in Latin America?
Thanks in advance for your comments.
If you only use your own training data to build an NLU model through the DIETClassifier, then you’ll have the exact same experience in Spanish as in any other language - it all depends on the quality of your training data.
Alternatively, you can use already pre-trained language models e.g. through Spacy - here, there’s generally better english models going around than for any other language.
More details about Rasa NLU in the docs