Quick Start

Build your first voice AI agent and talk to it in your browser — in about 5 minutes. Everything happens in the InfraVoice dashboard at your-infravoice-host; there's nothing to install.

What you'll need: a browser, a microphone, and your own provider keys for an LLM, Speech-to-Text (ASR), and Text-to-Speech (TTS). InfraVoice hosts no models — you bring your own keys, and new accounts start with free credit to try things out.

Step 1 · Create your account

Open your-infravoice-host and sign up with your first and last name, a work email, and a password (at least 8 characters). Press Create account — we'll email you a 6-digit verification code. Type it on the next screen to confirm your email, and you're in.

The very first time you land on your dashboard, a short welcome wizard walks you through the essentials. You can Skip for now and reopen it later from Getting started.

Step 2 · Meet your demo agents

You don't start from a blank screen. Every new workspace comes pre-loaded with three working demo agents so you can explore right away. Open Agents in the left sidebar to see them — each row shows its status and call count, and an Active badge means it's published and ready to take calls.

You can open a demo agent and talk to it immediately (skip to Step 5), or build your own from scratch in the next step.

Step 3 · Create or open an agent

Think of each agent as one "employee" with a specific job — a support rep, a scheduler, a salesperson. Click Create an Agent (or open an existing one) to reach the editor. On the Configuration tab:

1. Agent Name — something memorable, e.g. "Front-desk Booking Bot".
2. System Prompt — the most important field. Describe who the agent is and how it should behave, in plain English: "You are Emma, a warm AI assistant for City Clinic. Help patients book, reschedule, or cancel appointments. Keep replies short, and always confirm the date and time before finishing."
3. Welcome Message — the first thing the agent says when a call connects.
4. Pick a Language (leave the other defaults if unsure).

Click Save Changes.

Tip: Put behaviour ("always greet by name, never argue") in the system prompt, and save facts ("our hours are 9–5") for the Knowledge Base tab.

Step 4 · Add the three providers

An agent needs three engines before it can talk. Open the Providers tab and set each one:

For each engine, pick the provider, paste your API key, choose a model (and a voice for TTS), then save with Save LLM, Save ASR, and Save TTS.

All three are required before an agent can take a live call. To avoid re-typing keys on every agent, save them once under Build → Credentials and reuse them everywhere — keys are encrypted at rest and never shown in full.

Step 5 · Talk to it — Voice Test

The fastest way to hear your agent. Open Voice Test (the Voice Test Studio) from the sidebar:

1. Choose your agent and click Start Call.
2. Allow microphone access when your browser asks.
3. Speak naturally — the agent listens, thinks, and replies out loud. Watch the live transcript.
4. Use Mute if needed, and End Call when you're done.

That's a real, live voice conversation right in your browser — no phone required.

Note: Use a headset for the cleanest experience. Live voice calls draw a small amount of wallet credit (billed by the second); building and editing agents is free.

Step 6 · Go live on a phone number

Happy with it? Take it live in two moves:

1. Publish. Saving edits creates a draft — calls keep using the last published version until you push the new one. Use the version banner at the top of the editor: a new agent shows Not published yet → click Publish v1. When everything is live, the banner reads Live.
2. Attach a phone number. Open the agent's Telephony tab, enter your Twilio Account SID, Auth Token, and phone number (in E.164, e.g. +14155551234), and click Save Telephony. Then copy the read-only Webhook URL and paste it into your Twilio number's "A call comes in" setting (Webhook, HTTP POST). Dial the number and your agent answers.

The User Guide walks through the full Twilio setup step by step, including handing live calls to human agents.

Next steps

You've built, tested, and published your first agent. From here you can:

• User Guide → — the full end-to-end walkthrough of every feature.
• Add a Knowledge Base so the agent answers from your own FAQs and policies.
• Add Tools so it can do things mid-call — look up an order, book a slot.
• Set up a Call Center to hand live calls to human agents and watch them in real time.
• Use QA, Call History, and Analytics to score and improve every call.