Voice AI

I have a client who would like his inbound AI employee to use his own voice so that his customer interactions are more personable. I would like to be able to upload an AI voice file that I've trained with external software, or be able to train the custom AI voice within GHL. Thank you for all you do for us! ie. https://elevenlabs.io/voice-cloning

An ability to test or demonstrate the Voice AI by just clicking a widget or a button in a funnel/website.

With companies like Nedzo that have ultra realistic voices, I think the HL voice ai has a long way to go. Would be good if the voices could sound less robotic.

This would be great for 2 reasons. 1) When one of your contacts is calling using the AI phone number they would be recognized " Hi John" Thanks for calling, how can I help you? 2) When doing voicemail drops to sell AI voice the caller could call back, be recognized and have a demo of what their missing on their call back.

Best

The first sentence the Voice AI speaks usually determines whether or not the caller will know it's AI and hang up, or feel comfortable continuing the conversation. While it is currently a feature to use your own voice for AI receptionists, they don't sound great if used for the entire conversation. Thus it'd be great to be able to personally record just the welcome message so that it sounds human, then the rest of the conversation is carried on by the AI voice. It will encourage callers to state the reason that they're calling before hanging up, so that the AI agent can at least be given the chance to respond to their needs.

AI Studio already has the ability to build blogs, but today I don’t feel comfortable using it for production websites because the content layer isn’t dependable enough. I’ve personally experienced blog articles being modified or completely disappearing after prompting AI Studio to make changes that had nothing to do with the blog itself. That makes it difficult to trust AI Studio for SEO-focused websites where articles represent months or years of work. Feature Request Please separate content from page generation by introducing a dedicated CMS layer inside AI Studio. Instead of storing blog articles as part of the AI-generated page structure, store them in a protected content database that AI prompts cannot accidentally overwrite. The CMS should include: * Blog Posts * Categories * Tags * Authors * Media Library * Draft / Published states * Scheduled publishing * Revision history * Autosave * Restore previous versions * Backups * Dynamic routing (/blog/[slug]) * Search * RSS Feed * Sitemap generation * SEO fields (Title, Description, Open Graph, Canonical) Most Important Requirement AI prompts should never modify or delete CMS content unless the user explicitly asks to edit that specific article. For example: If I prompt AI Studio: “Redesign the homepage.” or “Improve the Services page.” or “Change the color palette.” My existing blog articles should remain completely untouched. Likewise, if I ask AI Studio to generate a new page or redesign another section of the website, it should never rewrite, regenerate, move, or delete existing blog content unless I explicitly instruct it to do so. Why this matters Many agencies and businesses rely on blogs for long-term SEO. Articles often represent dozens or hundreds of hours of work and become valuable business assets over time. Right now, AI Studio is an amazing website builder, but because blog content can be unintentionally affected by unrelated prompts, it’s difficult to trust it as the primary platform for content-driven websites. A protected CMS architecture would make AI Studio a true replacement for WordPress, Webflow CMS, and other content management systems while giving users confidence that their content is safe.

As we work towards offering AI receptionists to clients, it would be super helpful if the AI could train based on previous call transcripts. For example if we have someone call in and ask to speak with Patty, the AI may not realize they're asking for Patricia. Or maybe industry jargon is used on calls, it would be helpful if the AI can learn what all those terms are so it can respond properly. Some of this stuff only turns up in client calls, and in an ideal scenario, after the call the AI would ask the sub account owner or admin to clarify or add context to what was discussed on the call.

Feature Request: Please extend the existing DTMF (keypad button press) capability — currently only available through the "transfer_call" action — so it can also be used as a standalone action during outbound calls, allowing the AI Voice Agent to navigate automated phone menus (IVRs) when calling businesses. Use Case: When doing outbound B2B calling — for example, calling a business's main line — the call almost always hits an automated IVR system first ("Press 1 for Sales, Press 2 for Support," etc.) before reaching a live person. GHL's Voice AI Agent already has DTMF functionality built for the transfer_call action, but it isn't available earlier in the call flow. Since the agent has no way to press digits before that point, it gets stuck at the IVR and can't get through to a live rep. Right now every call that hits an IVR just stalls or drops. What's needed: Ability for the AI agent to detect/listen for IVR prompts and either: Automatically press a pre-configured digit sequence at a defined point in the call, or Dynamically decide which digit to press based on the IVR prompt it hears (ideal), or At minimum, decouple the existing "press digit" tool from transfer_call so it can be added as its own standalone action the agent can trigger mid-call, any time during an outbound call Why this matters: This is already a standard, supported feature with every major AI voice platform — Vapi, ElevenLabs (Conversational AI), and Retell AI all support DTMF/keypad press actions as an independent tool available throughout the call, not gated behind a transfer action. GHL already has the underlying DTMF capability built — it's just tied to transfer_call — so this is a real blocker for anyone trying to use it for serious outbound B2B calling where reaching a live rep through a phone tree is the norm, not the exception. Adding this would bring GHL's voice agent to full feature parity with the other platforms and unlock a huge use case (outbound dialing to businesses) that's currently not viable with the current setup.

Would be extremely helpful to keep the same point of contact.