Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.ravan.ai/llms.txt

Use this file to discover all available pages before exploring further.

The Agents dashboard is where you design the core intelligence of your voice infrastructure. From this central view, you can monitor the real-time performance of every agent in your organization.
Main Agents Dashboard showing active and inactive agents
Each agent card provides a snapshot of activity, including campaign success rates and call statistics (Total, Done, Outbound, Inbound, and Web).

​
Quick Start

Get your first agent live in under 5 minutes:

Click + New Agent

From the Agents dashboard, click + New Agent to open the agent builder.

Write a system prompt

Describe who the agent is and how it should handle calls.

Pick a voice and model

Select your AI model and voice profile (like Iris).

Set a welcome message

Define how the AI begins the conversation.

Test with a web call

Use the testing panel to have a live conversation immediately.

​
Creating a New Agent

When you enter the agent builder, you have a unified workspace to manage the prompt on the left and technical configurations on the right.
The full Agni agent builder interface

​
AI Provider Settings

Configure the underlying reasoning engine and the “personality” of the voice:
  • Model: Choose between Thunder Emotion for high-fidelity reasoning or Thunder Emotion Lite for faster responses.
AI Model selection dropdown
  • Voice: Access our library of 28+ voices. Iris is our most popular female voice, known for its natural cadence and professional tone.
Voice selection gallery with Iris highlighted

​
Welcome Message

Under the Welcome Message tab, you can decide how the interaction starts. If the AI speaks first, you can choose a static custom message or a Dynamic message based on prompt, which allows the AI to tailor its greeting to the specific context of the call.
Options for setting dynamic or custom welcome messages

​
Agent Configuration

​
Functions & Tools

Equip your agent with the ability to take action during a call.
Dropdown menu for agent functions like End Call and Transfer
  • IVR / Press Digit: Perfect for navigating automated phone menus. You can set the Pause Detection Delay to ensure the agent waits for the right moment to press a digit.
  • Custom Functions: Connect your own backend.
    • Custom (Server-Side): Set up a direct API call (GET, POST, etc.) with specific headers and JSON schemas.
    • Client Function: Define parameters that the LLM will pass back to your local client application.
The configuration screen for a server-side custom function

​
Specialized Settings

Fine-tune the audio environment. Add Background Sounds like an office hum or city noise. You can also adjust Interruption Sensitivity and set Reminder Messages to prompt the user if they stay silent for too long.
Control telephony rules like Voicemail Detection and Max Duration. Enable Memory to allow the agent to recognize and remember context from returning callers.
Define exactly what data to pull from the transcript. This is ideal for capturing Sentiments (Positive/Negative) or generating a Summary to pass to a human agent during a transfer.
Link a specific knowledge base so your agent can answer questions using your company’s internal documentation and FAQs.

​
Testing Your Agent

The built-in testing environment allows you to “break” your agent in a safe space before going live.
The testing sidebar for Web and Phone calls

​
Test Modalities

  • Web Call: Talk to your agent directly through your browser. You can also inject Metadata (JSON) to test how the agent handles dynamic prompt variables.
Using Dynamic Variables: To set up dynamic variables, simply include them in your Agent Prompt using double curly braces (e.g., {{user_name}} or {{company_name}}). Once added, new input fields will automatically appear in your testing panel, allowing you to pass specific values into these variables during the test.
Testing panel showing dynamic variable inputs
  • Phone Call: Enter your phone number to receive a live call from the agent. This allows you to experience exactly how the voice interactions sound over a real cellular network with standard latency.
Pro Tip: After completing a test call, always review the Post-Call Data section. This helps verify that the AI is accurately extracting the intended summaries, sentiments, and data points.