Design
Tools for designing different types of assets.
The main editing environment for conversation design. This is where you create, test, and modify your dialogues – the node graphs as well as the related code.
Read this article for more details.
Create, train, and modify entities and entity datasets.
An entity refers to a specific object, concept, or element in a text that is recognized and classified based on its relevance to a particular domain or task. Entities can be anything from people, locations, and organizations to dates, times, and numerical values.
More about entities here.
Create and modify reusable dialogue graph snippets. Snippets can save you time as you won't need to create the same structure again and again – you can just easily drag&drop a copy of a pre-made snippet into any dialogue.
Read this article for more details.
Create and modify dialogue mixins. These are referenceable contents of different types of nodes (Intent, Speech, Command, Function, User Input, or the "Init" code of a dialogue).
You can link a mixin to a specific node, and any updates made to the mixin content will automatically update the node (after the next dialogue build).
Read this article for more details.
Upload file assets (images, sounds, JSON files) that you will be able to reference and use in your dialogues.
In case you need a direct URL of a file asset, you will find it in the Properties of the file in question.
Design digital personas that will handle your dialogues. A persona is defined by a combination of certain features like voice, gender, or the implicit dialogue containing persona-specific reactions to relevant global intents.
Reference existing TTS (text-to-speech) voices from external providers. These voices can then be selected in Personas or Dialogues.
Currently supported providers: Amazon, Google, Microsoft.
If you are creating a new reference:
in TTS Configuration — Provider, type the name of the provider ("Amazon" or "Google" or "Microsoft")
in TTS Configuration — Locale, choose the corresponding locale specified by the TTS provider
in TTS Configuration — Name, type the official name of the voice specified by the TTS provider
Reference existing ASR (automatic speech recognition) settings from external providers. These speech recognizers can then be selected in Personas or Dialogues.
Currently supported providers: Amazon, Google, Microsoft, Deepgram.
If you are creating a new reference, make sure all settings are compatible with your selected provider.
Last updated