Our small talk API is trained with over 100M data sets, for understanding natural language. Thanks to our virtual AI assistant which assists over 3.5M users worldwide, we've collected almost every intent when conversing in small talk, with a unique accuracy of over 91% for most valid use cases. Today we are officially releasing our trained AI for intent/entity extraction to the public.
The model understands over 40 intents! For example:
And much more!
Contact us at firstname.lastname@example.org for any questions.