Skip to main content

​
Quickstart

Follow these steps to create, configure, and test your first AI voice agent on Callem Studio.

​
Prerequisites

  • A Callem Studio account — sign up or ask your team admin for an invite
  • A project created in the dashboard

​
Step 1: Create an Agent

1

Navigate to Agents

Open the Build > Agents section from the sidebar.
2

Create a new agent

Click New Agent and give it a name (e.g. “Reception Assistant”).
3

Configure the basics

  • System prompt: describe your agent’s role, personality, and instructions
  • Voice: pick a TTS voice from the available providers (ElevenLabs, Azure, etc.)
  • Language: select the primary language for your agent
  • LLM model: choose the underlying language model

​
Step 2: Add Knowledge

1

Navigate to Knowledge Base

Open Build > Knowledge Base from the sidebar.
2

Create a knowledge item

Click Add and paste or upload the content your agent should know — FAQs, product descriptions, policies, etc.
3

Link to your agent

In your agent’s configuration, select the knowledge base to attach.

​
Step 3: Assign a Phone Number

1

Navigate to Channels

Open Deploy > Channels from the sidebar.
2

Configure a number

Select an available phone number (or import your own via SIP) and assign your agent to it.

​
Step 4: Make a Test Call

Call the assigned phone number from your own phone. Your AI agent will answer and respond based on the prompt and knowledge base you configured.

​
Step 5: Review the Call

1

Navigate to Calls

Open Monitor > Calls from the sidebar.
2

Inspect the call

Find your test call in the list. Click on it to see the full transcript, listen to the recording, and review the call analysis.

​
Next Steps

Add Tools

Let your agent call external APIs during conversations.

Build a Workflow

Create multi-step conversation flows with branching logic.

Launch a Campaign

Automate outbound calls to a list of contacts.

Set Up Alerts

Get notified when call metrics cross thresholds.