Listen. Check. Resolve.
Voses replies to the people who message your business using only what you've published, checks every request before it answers on its own, and hands anything sensitive to the right person on your team with the full context.
Illustrative example with sample people, businesses and prices.
The AI model drafts. Voses decides what is allowed to happen. That separation is the whole product, and it's why Voses can be trusted with a clinic's patients or a law firm's clients.
Every message lands in one record for that customer, with their history and which business they're talking to. Nobody repeats themselves, whether the AI or your team replies.
Before the AI replies on its own, Voses checks what is being asked. It only answers straightforward questions from your information. Complaints, requests for a person and anything it isn't confident about go straight to your team.
Simple questions are answered from your knowledge base, with the source shown to your team. Everything else is handed over with a summary and a suggested next step, and all of it is logged.
A full skin check with one of our GPs is $95.
Start with auto-reply off. Your staff ask the AI for a draft, check it and send it themselves. When you're happy with the drafts, switch auto-reply on. Even then, Voses only answers straightforward questions on its own and hands everything else to your team.
Hi, I'm the Harbour Medical assistant. I'm an AI, and you can ask for a person at any time.
Clinics, law firms, accountants and IT providers can't afford a chatbot that improvises. These are design decisions, not settings. Read how we handle security and data.
If the answer isn't in what you've published, Voses doesn't guess. The question goes to your team.
Complaints, requests for a person and anything Voses isn't confident about are handed to your team. A wrong handoff costs one staff reply. A wrong answer can cost much more.
Every record is scoped to the business it belongs to. No other account can see your customers or conversations.
Every message shows whether the AI or a person sent it, AI spending is tracked for each business, and deletions are recorded with who made them and when.
We ship carefully and say so plainly. This is the current state of the product.
| Capability | Status |
|---|---|
| Website chatOne line of code on your site | Available |
| Team inbox and handoffStaff take over with the full history | Available |
| Knowledge baseThe articles and facts Voses answers from | Available |
| Phone callsThe same rules on your business number | In development |
| Email and SMSMore channels, one customer record | In development |
| Calendar bookingBooks into the calendar you already use | In development |
We're setting Voses up with a small group of businesses first, and it costs nothing while we do. Before any paid plan starts we'll publish pricing and give you at least 30 days' notice. You can stop using Voses at any time.
$0while we onboard our first businesses
I'm Mark Julius Gimena. My background is in managed IT and cybersecurity for small businesses, and I spend my days supporting the kind of clinics, firms and offices Voses is built for. Teams of a handful of people, answering the same questions over and over, where one wrong answer to a customer can cause a real problem.
I built Voses to be the assistant I wanted those businesses to have. It answers from their own information, follows their rules, and steps back the moment something needs a person. Early access means you'll be working with me directly, and your feedback shapes what gets built next.
Mark Julius GimenaFounder, Voses
Can't see yours? Ask in the enquiry form below and we'll answer it personally.
Voses only answers from the information you've published, and checks what's being asked before it replies on its own. Anything outside that goes to your team instead of a guess. We recommend starting with auto-reply off, so your staff check AI drafts before sending them, and switching auto-reply on once you're comfortable. Even with auto-reply on, Voses only answers straightforward questions by itself. Your team can take over any conversation.
Yes, and that's the point. Voses is upfront that it's an AI assistant, and customers can ask for a person at any time. Being honest about it doesn't mean sounding like a robot, though. Voses talks the way a good front desk person would: warm, natural and straight to the point, using your business's own information. No scripted menus, no stiff bot replies, just a real conversation. Your staff can also see which replies came from the AI in the inbox.
Yes. You can switch AI auto-reply off at any time, and your staff can still ask it for drafts to check and send themselves. You can also turn the website chat off completely.
Your data is stored in a managed Postgres database hosted by Supabase, in [REGION TO CONFIRM]. Each business's records are kept separate from every other business. The details are on our security and data page.
No. Voses doesn't train AI models. We use OpenAI's API to write replies, with response storage turned off, and under OpenAI's API terms that data isn't used to train their models.
Your key business information (prices, hours, policies), a link to your privacy policy, and the ability to add one line of code to your website. During early access we'll help you set all of it up.
Nothing during early access. We'll publish pricing and give you at least 30 days' notice before any paid plan starts.
Tell us a little about your business and what your customers usually ask. We'll reply within one business day to set up a short call.
Prefer email? welcome@voses.app