Skip to main content

​
Business Mode & Developer Mode

Agent Studio comes in two views. You switch between them with a single toggle in the top bar.
  • Business Mode — simplified workspace. Only the settings that matter for getting a voice agent live: Behavior, Voice, Knowledge, Tools, Call, Activity, Deploy. Industry-standard defaults are applied silently so a non-technical user can create an agent and hit Try Agent in under 30 seconds.
  • Developer Mode — the full studio. Every panel, every advanced knob — model temperature, structured output, hooks, guardrails, evaluations, A2A, flow editor, the lot.
The mode is purely a rendering choice. Switching never destroys settings; Business Mode just hides the advanced ones.

​
Which view will I see?

YouDefault view
Existing account (pre-launch user)Developer Mode — the full studio you’ve always used. No change.
New accountBusiness Mode — the simplified view. Flip the toggle anytime to see the full studio.

​
The toggle

In the studio top bar, between the cost breakdown and the Save button, you’ll see a Developer Mode pill with a small rotating border.
  • Click the pill to flip your account-wide default. The change applies to every agent you open from now on.
  • Click the three-dot menu next to the pill to pin this one agent to a specific mode, independent of your default. Choose:
    • Follow default — agent inherits whatever your account default is set to.
    • Pin to Business Mode — this agent always opens in Business Mode, even if your default is Developer.
    • Pin to Developer Mode — this agent always opens in Developer Mode, even if your default is Business.
When an agent is pinned, a small lock icon appears on the pill so you can tell at a glance whether you’re seeing the agent’s view or your default.

​
What Business Mode shows

Left rail — 8 items instead of 15:
TabWhat’s in it
OverviewAgent name, model summary, cost-per-minute, “Try Agent” shortcuts.
BehaviorGreeting + Core Behavior + Variables. A one-click ✨ Enhance with AI button rewrites a section using a strong model and shows the suggested version side-by-side with the current text. Accept or reject.
VoiceLanguages (multi-select chips) + voice grid (paginated, searchable, filterable by gender / language / provider) + speed slider. Your own cloned voices appear first, then Aero, then other providers.
KnowledgeAttach knowledge base sources.
ToolsThe full tool library, grouped into Pre-call / In-call / Post-call so it’s clear when each tool runs.
CallMax call duration, idle timeout, noise filtering, voicemail. The four knobs that matter for a phone call.
ActivityRead-only log of recent calls / sessions.
DeployChannels, phone-number assignment.
Hidden in Business Mode (visible in Developer Mode): Flow editor, Performance, Skills, Learning, Evaluations, Approvals, A2A, Versions, embed-widget config, custom analysis, event subscription, A/B testing, plus every advanced knob inside the panels above (temperature, structured-output schema, reasoning effort, HITL, hooks, guardrails, S2S turn detection, custom STT/TTS provider overrides, etc.).

​
What gets applied automatically

When a new agent is created from the Create Agent flow, the platform ships it with a production-grade baseline so Try Agent (Web / Phone / Chat) works on the first click — no configuration required.
  • TTS — Aero (our own low-latency engine), default voice Riya, with Sarvam as the automatic fallback if Aero is briefly unreachable.
  • STT — picked automatically from the languages you select in the Create Agent picker: Sarvam for Indian languages (Hindi, Marathi, Bengali, Tamil, Telugu, Kannada, Gujarati, Malayalam, Punjabi, Urdu, Odia, Assamese, Nepali), Deepgram for English and everything else.
  • Languages — English (India) + Hindi by default. You can add or remove from the Voice tab; the speech-recognition engine swaps automatically.
  • Greeting — "Hi {{caller_name}}, how can I help you today?". The {{caller_name}} variable is filled from your CRM lead data when available, and dropped gracefully otherwise.
  • Call lifecycle — 4-minute test-call cap, 10-second silence timeout. Voicemail detection, call recording, DTMF are off by default but visible in Business Mode if you want to flip them on.
  • Conversational tuning — interruption handling on, noise filtering at the production-default strength, turn detection in auto mode.
In Developer Mode you can override every one of these from the relevant panel.

​
Switching modes mid-edit

You won’t lose anything. If you set, say, model temperature to 0.2 in Developer Mode and then flip to Business Mode, the value stays on the agent — Business Mode just hides the control. Flip back and it’s still 0.2. If you’re on a tab that doesn’t exist in the view you switch to (Flow editor, for example), the studio snaps you back to the Behavior tab so you’re never staring at an empty pane.

​
Choosing for your team

  • Solo founder / non-technical user who just wants a voice agent that works — leave the default on Business Mode. The simplified studio + smart defaults are designed for this.
  • Developer / agency configuring many agents with custom logic — flip your default to Developer Mode in the top bar. You’ll see the full studio on every agent you open.
  • Mixed team — leave your account default wherever you prefer, and use the per-agent pin for the agents that need the other view. A non-technical user can pin “their” agent to Business Mode without affecting how a developer on the same team sees their own agents.