Configure
Everything you set up before building conversation logic – engines, personas, knowledge, and profiles. A map of the Virbe Dashboard.
Before You Build
Configuration happens in the Virbe Dashboard, and everything you set up here is shared across every Profile and touchpoint you later deploy – get it right once, and every deployment inherits it.
Dashboard navigation
The Virbe Dashboard is organized into eight sections via the left sidebar:
| Section | What you do here |
|---|---|
| Overview | See dashboard status, manage Profiles |
| Configurations | Connect speech and AI engines |
| Personas | Define avatar, voice, and identity |
| Knowledge Base | Add documents for AI-powered answers |
| Conversation Editor | Build the Pipeline (covered in Build) |
| Conversations | Review conversation history (covered in Monitor) |
| Analytics | View usage metrics (covered in Monitor) |
| Issues | See system warnings (covered in Monitor) |
What to configure and in what order
If you're setting up a Dashboard for the first time, work through these in order:
1. Configurations – Connect the engines that power the virtual being's voice and intelligence. You need at minimum one STT engine, one TTS engine, and one AI Model before a conversation can work end to end.
2. Personas – Create the identity and appearance of your virtual being.
3. Knowledge Base – Add any documents or collections the assistant should be able to draw on when answering questions. Optional, but needed if you're using the Tool Agent or LLM Response nodes with RAG.
4. Profiles – Create at least one Profile to define a deployment target. Profiles tie everything together: Persona, engine, speech settings, and touchpoint (web or kiosk).
In this section
Profiles
Deployment configurations that tie together a persona, engine, and touchpoint.
Personas
Avatar, voice, and identity of your virtual being.
Knowledge Base
Documents and collections for AI-powered answers.
Conversational Engines
Connect OpenAI, Dify, Dialogflow, or your own backend.
Speech & Language
Speech-to-text, text-to-speech, and AI model settings.