Questions, answered plainly

No. The assistant sounds natural and friendly, but it never claims or implies to be human. If anyone asks whether they're talking to a person, it says plainly that it's an AI assistant and offers to get someone from your team.

By default the greeting says so up front too: if your greeting doesn't already mention that it's a virtual assistant, "I'm an AI assistant" is added to it. Several US states require this, and it's simply honest.

Conversations run on Anthropic's Claude models. It answers from your business profile, FAQ, documents and connected software, and won't quote a price, discount or appointment time that isn't in your information or your calendar; it says someone will confirm instead.

It transfers to the right person if someone is available, or takes a detailed message and creates a follow-up. Questions it couldn't answer are listed for you, so you can add the answer once.

The caller hears a short recovery line and the call carries on; if the problem continues, the call moves to a transfer or voicemail rather than going silent.

Card numbers and Social Security numbers are replaced before anything is stored, shown on the live monitor, summarised or sent to the AI. Your software logins, voicemails and caller memory are encrypted at rest. A caller can only ever reach their own records, and only after the identity check you choose.

Yes. One click exports or erases everything about a caller, for GDPR and CCPA requests. You can also set how many days calls are kept, delete single conversations, and turn call recording off.

A strict health-privacy mode keeps patient details out of emails, texts and phone alerts, and the assistant never gives medical advice. Talk to us about your own compliance requirements before you go live.

No. Get a number in the dashboard and forward your existing line to it, always or only when you're busy or don't answer. To have it answer texts too, use the new number itself or move your number over.

Most of an afternoon: pick an industry template, fill in your hours, FAQ and team, connect your software, and test it by chatting with it or having it call your phone. Setup can also read your website and draft your profile for you to review.

Plans are priced by call volume and every plan has every feature. See pricing for the current plans and the free trial.

No. Spam calls and robocalls, blocked callers, spoofed numbers sent to voicemail, hang-ups before a conversation and any call under 15 seconds don't use your minutes, and spam texts don't count toward your texts. Your usage page shows how many calls were free each month.

Yes: your name, logo and colour on the dashboard, emails and your own domain, a portfolio view of every client, and billing reports per client. Get in touch.

The product is new, so here is exactly where it is. Every integration is built from the vendor's published API and tested against a simulated version of that API. None has been run against a live vendor account yet, and the phone, texting and AI services haven't yet carried live customer traffic.

That means the first live setup for your business is also the first real check of your integrations. We'll go through it with you, run practice calls, and test each connection before your customers rely on it.

Twilio, fully: calls, texts, recordings and transfers (not yet run against a live account). Telnyx numbers can be used for texts and voicemail.

Our own AI assistant, the same product you'd get, set up with information about AI Receptionist. It can answer questions and take a message for the team. It can't book anything or see anyone's account.