Skip to main content
You’ll build a working voice agent without writing any code: create it, write its system prompt, preview the conversation, and publish a version.
1

Create an agent

In the dashboard, open ZeptarAgents → Agents and choose Create agent. Give it a name. It starts on its Main branch with a first version.
2

Write the system prompt

The system prompt is where almost all of an agent’s behavior comes from — its persona, goal, tone, and guardrails. Keep it concrete and write for the ear, since the agent speaks its replies. For example:
You are a friendly front-desk assistant for a dental clinic. Greet callers, answer questions about hours and location, and offer to take a message. Keep replies to one or two sentences.
Edits autosave to your draft on the current branch — nothing ships until you publish. See System prompt.
3

Preview it

Open the live preview and actually talk to the agent. Iterate on the prompt until it behaves the way you want.
4

Publish or branch

Publish a version on Main when it’s ready, or fork a branch to try a riskier change in isolation and merge it back later. Every publish is an immutable version you can roll back to. See Versioning.

Next steps

How agents work

The speech pipeline and what you configure today.

Automate with the API

Create and manage agents from your own backend.