What's new in v0.1.12

Every agent now has its own voice link: a talk page where anyone can speak with the agent out loud, in its own voice. Open an agent and the new Voice link panel shows the link with Copy, an on/off switch and your plan's voice minutes. There's nothing to set up, and the link always follows the agent: change its voice or name and the link uses the new one.

Voice links work on any plan that includes voice minutes. On a plan without them the panel shows the upgrade note, and the link starts working as soon as you upgrade. A visitor who opens a talk page while voice isn't available sees a short message, plus your booking or chat link if you've set one, before being asked for their microphone.

Voice links also appear under Voice deployments, marked Voice link, where you can adjust their limits, website addresses and disclosure. The agent and voice are read-only there because they follow the agent.

The setup wizard's Meet your agents step now has More voices… under each agent. It opens the full voice library, with voices matching the agent's gender listed first, and you can hear each one before you pick it. You can switch back to the recommended voice at any time. More voices is on Team and above; on other plans it's shown locked with a link to the plans, and the three recommended voices stay one click away on every plan.

Voice samples ("hear this voice") now play instantly when you replay a voice and phrase you've already heard. New samples are limited to 100 a day per workspace, on every plan; if you reach the limit, the voice picker tells you, and new samples are available again the next day.

Test connection now works for demo targets that sign in as another Appricity workspace: it signs in exactly the way a demo does, and says what to check when it can't. The edit page also shows that target's workspace id instead of an empty box, so it no longer looks as if saving lost it.

Settings now shows your Workspace ID with a Copy button, which is what you enter when another workspace's demo target signs in as this one.

Creating a meeting type no longer ticks off the "Go live" setup step. That step now completes only when something a buyer can actually reach is live: a website embed, a voice page, or a channel your agent answers on.

Role-surface dropdowns no longer list Generalist, BDR, Sales Engineer, Customer Success and Product Trainer twice. Where your workspace has its own copy of a role-surface, that's the one offered; one shared from a parent workspace with no copy of your own is still listed, marked "inherited". The Role-surfaces page still shows both.

Your public voice page now introduces your agent and your company, not ours. The heading, the on-screen and spoken AI disclosure, the status line and the transcript all use your agent's name and your company name (your brand pack's name when the voice surface uses one), and the page shows your brand pack's logo. When no name is set yet, the page and the disclosure stay neutral ("You are speaking with an AI assistant — not a person.") rather than naming anyone. New voice surfaces now start from that neutral disclosure, which you can edit.

Creating a voice deployment now lets you pick its voice from a searchable list — name, accent, gender and use case, with a ▶ Play button to hear it — instead of typing a voice id by hand. The list follows the provider you choose (Kokoro or ElevenLabs; with your own ElevenLabs key you see your account's voices). The form also starts on a voice that works where you are, rather than one that may be silent.