Natural Language Understanding
that Performs with Less Data

Our NLU is what we categorize as a few-shots. It doesn't require a lot of data, sometimes 10 examples for an intent will be enough. It has a direct impact on how fast it trains, but more importantly how fast you can put it in the hands of actual users. This is a huge barrier to entry for developers starting out. If you need 100 utterances per intent just to get started, it might be difficult to come up with a solid proof of concept that you can build on top of. With our platform, you just get it done faster.

Botpress supports 12 languages out of the box. English, French, Spanish, Japanese and Arabic are only a few examples of the languages we support natively. If you need a language that is not built-in, we give you the flexibility to use a FastText model which is available in 157 different languages. With Botpress Enterprise, you can also support as many languages as you need simultaneously without duplicating the work. You don't need multiple chatbots to create a multi-language conversational experience. We have built-in translation to avoid doing the tedious task of duplicating your work.

The context is essentially the memory of past messages or interactions. By leveraging topics and intents created by the chatbot developer, your chatbot can create an interaction that is much more fluid and natural. It will be capable of handling slot filling and context switching by default to create an outstanding conversational experience for your customers and your teams.