- What does the AI talk about?
- It follows the conversation setup for each use case: its role and tone (for example, a calm dental clinic receptionist), the opening greeting, the details to collect, answers to common questions, and the conditions for handing off to staff. You set these in the dashboard or through the API.
- Which languages can it talk in?
- Japanese, English and other languages.
- Where are call audio and conversations processed?
- Speech recognition, generative AI and speech synthesis combine APIs from inside and outside Japan, so call audio and conversation content may be processed outside Japan.
- Does voicast provide phone numbers?
- No. You use the numbers and lines you already have (for example, your SaaS’s Twilio account). voicast receives the call audio over a WebSocket and holds the conversation.
- What happens to calls if voicast is down?
- If it cannot be reached or the connection key is wrong, the stream closes immediately and the call moves on to the next step in your flow (for example, ringing your staff).
- Can I change which details are collected?
- Yes. Besides name, callback number and reason for the call, you can add your own fields to the conversation setup. Fields can be text or phone numbers.
- What happens to live calls when I change the setup?
- Setups are versioned. A call in progress keeps the version it started with, and the next call uses the new one.
- How do I receive the result?
- After the stream closes, call
GET /v1/calls/by-provider/{CallSid} with the Twilio CallSid. You get the outcome (completed / answered / handoff / hangup / error), reason, collected fields, duration and cost.
- How is pricing calculated?
- Pay as you go, per minute of calls. Contact us for details.