How it works
An overview of how you can get started designing with VoiceXD
Setup the design space for you and your team
Create a workspace
Create a new workspace to host your team's projects. You can use our always free Starter plan or upgrade to our Team plan for unlimited projects and editors. See our plans here.
Add your team
Invite all of your team members to the workspace and set roles for owners, editors, and viewers (who can still leave comments).
Organize your assistants
Create a new project to start designing. If you have multiple assistants, you can store them within a single project or create multiple different ones. You can design for any modality within a single project.
Begin writing initial conversation paths for each of your assistant's use cases
Write user inputs without intents
Design user inputs by writing out utterances and training phrases and creating buttons
Design the assistant responses
Create responses with variations, no-matches, no inputs, and images
Work in the script or model view
Design in the script view which shows a single path at a time or the model view where you can view all paths
Capture user data
Use dialogue variables to track and manage user provided information
Design scripts with AI-Companion
Use AI to write conversation paths for each of your assistant's use cases
Build entire conversation models just by using AI generation
Auto-generate entire conversation paths with variables just by describing the goal of the path.
Generate content from phrases
Use sample user utterances and assistant responses to generate training phrases, response variations, no-matches, and no-inputs
Auto-populate variable values
Use AI-Companion to generate values and synonyms for both dialogue and context variables
Customize the AI-generated content
Choose traits that match your assistant's personality and brand voice to generate content that reflects it
Build the conversation model and scale your content
Add more complexity to your designs by modeling different scenarios, recovery paths, and more.
Create intents and response components
Attach your individual user steps to intents and convert responses that are used frequently into reusable components. Changes made to any intent or component are propagated to all instances.
Model any kind of context
Use context variables to simulate how any type of information external to the dialogue can influence the conversation
Create new scenarios with conditions and branches
Design new flows branching off of your happy path based on variations in user input, different contexts, and to repair user or system errors
Enable script scoping and context-switching
Sort scripts into groups based on their topics and choose whether they can be accessed from any script (global) or only within the same group (local). Decide which global and local scripts can be accessed after a particular assistant step to control context-switching.
Test & Share
Experience your designs with the click of a button and share your models and prototype with others.
Share and collaborate
Edit and walk through designs in real-time with your team and leave comments anywhere in the model canvas or on a specific node.
Create testing personas to simulate how different types of users will experience interactions with your assistant.
Personalized welcome messages
Create unique welcome messages for different testing personas and interaction modes (chat or voice).
Present and test your experience with a specific testing persona using voice, chat, or a combination of both. The possibilities are endless.