AI SDR agents

Create and configure inbound AI sales agents for conversations, qualification, meetings, and CRM synchronization.

An AI SDR agent is the configured Clara persona that handles your inbound conversations. Each agent combines behavior fields, training material, qualification rules, booking destinations, and integrations - and each one can run on your website 24/7.

What an agent can do

Capability	Description
Conversations	Real-time chat, voice, or avatar sessions with visitors
Intent detection	Reads what visitors say and asks the right next question
Qualification	Evaluates leads against BANT, MEDDIC, CHAMP, or custom question sets
Presentations	Shows relevant slides, images, and videos during the conversation
Meeting booking	Schedules qualified buyers on your connected calendar
CRM synchronization	Creates and updates contacts, logs conversations, tracks meetings

Creating an agent

1. In your Clara workspace, click Create Agent.
2. Give Clara a name and fill in the behavior fields - Company Description, Goal, Tone, and Closing Behavior - to shape how she speaks and represents your brand.
3. Choose the avatar: pick a stock avatar from the gallery, upload a face, or generate one from a prompt.
4. Attach training: a knowledge base and visual presentations.
5. Save. The agent is live in your workspace and ready to embed.

Agent behavior controls

Control	What it does
Company Description	What your company offers, who it serves, and your key value proposition
Goal	What Clara drives the conversation toward
Tone	How Clara speaks
Closing Behavior	How Clara ends conversations and suggests next steps
Qualifying questions	The framework or custom set Clara asks during conversations
Suggested topics	Starter prompts shown in the widget
Allowed domains	Restrict which origins can embed the widget

Behavior fields

Four fields shape how Clara converses. Each accepts up to 1,000 characters:

• Company Description - describe your company, what it offers, who it serves, and your key value proposition.
• Goal - what the conversation should achieve: understand the visitor’s needs, qualify them, demonstrate the value of your platform, and guide the conversation toward a meaningful next step.
• Tone - how Clara speaks. For example: concise, friendly, confident, and professional, with simple language and no unnecessary technical jargon unless the visitor asks.
• Closing Behavior - how conversations end. For example: ask if the visitor has any final questions, suggest a clear next step such as booking a meeting or exploring a deeper demo, and end with a polite, encouraging call to action.

Behavior changes apply to the next conversation - active conversations finish on the previous config.

Next steps

Conversation workflows

Route and complete conversations with rules.

Templates

Standardize agent behavior across teams and sites.