More to configure: all seven panels

The collapsed section at the bottom of the agent page holds seven panels. What each one is for, and the order worth opening them in.

5 min readUpdated 26 August 2026

Below the main cards on the agent page sits More to configure, described on the page as voice and models, call handling, tools, data capture, workflow and embed, with an invitation to open any section to fine-tune the agent.

It looks like a small strip and it is not. There are seven panels behind it, and between them they hold most of what makes an agent good rather than merely working.

The seven, and what each is actually for

PanelThe page describes it asOpen it when
AI & voice modelsLLM, transcription and voiceThe default sounds wrong, or replies are slow
Call settingsGreeting, interruptions, end-callThe rhythm of the call feels off
Conversation intelligenceNatural speech, listening, summariesIt talks over people or sounds robotic
ToolsCalendar, webhooks, integrationsThe agent needs to do something, not just say something
Data captureCollect and save caller infoYou want structured records out of calls
WorkflowMulti-step conversation flowThe call genuinely has ordered steps
EmbedAdd the agent to your siteYou want it on your website, not only on the phone

You do not need any of them to start

A working agent needs a name, instructions, and the defaults. Everything in this section is an addition.

The useful habit is to open a panel because something is wrong, not because it is there. Listen to a call, notice the specific problem, then find the panel that owns it. Turning on settings speculatively is how people end up with an agent nobody can debug.

Tip

Change one thing, hear it with Talk to your agent, then change the next. Three changes at once and a call that still feels wrong tells you nothing.

Common questions

Do I have to configure all of these?
No. An agent works with none of them touched. They exist for specific problems and specific additions.
Where do I change the voice?
Either the Voice & model card higher up the page, or the AI & voice models panel here, which also covers transcription.
Which panel do I want if the agent interrupts people?
Call settings for interruption behaviour, and Conversation intelligence for how it listens and decides someone has finished.

Was this helpful?

Related