AI Agents
Customizable AI assistants with their own instructions: lead analysis, responses, classification.
An AI agent is an AI worker customized by you: it has a role, instructions, and works in automation scenarios.
Where to find
Automation Library module → AI Agents (…/ai-agents).
Creating an agent
- New agent → name and instructions (role, tone, rules: "You are a sales manager. Evaluate leads by budget and urgency...").
- Select an AI model (default is optimal).
- Use the agent as a node in scenarios: input data → agent's response → continue the flow.
Agent examples
- Lead qualifier — evaluates an application, returns a score and recommendation;
- First responder — drafts a response to a client (with human approval);
- Classifier — sorts inquiries by topic.
Agent operations consume AI credits.