Confidential by setup
We can configure the agent so conversations are used only to answer the customer and create support tickets. No medical advice, no unsafe answers, and unclear topics go to a human.
Client-facing voice agent demo
Click once, hear each voice, then choose the included-minute bundle that fits the client. Every route can run on a website now and can be connected to phone later.
Pick a route, hear it speak, then choose the package that fits your client.
The voice menu
Eight named voices, from cheapest demo mode to premium and private. Pick the voice, then pick 500, 1,000, 2,000, or request a quote for 2,000+ minutes.
Also from MyClienta
VoltType prepares the approved FAQs, support answers, sales lines, and follow-up messages. VoicePilot can then speak from that approved content, so the voice sounds useful instead of guessing.
For most clients, the important question is not the provider name. It is whether customer conversations stay confidential, who maintains the system, and what can be made private later.
We can configure the agent so conversations are used only to answer the customer and create support tickets. No medical advice, no unsafe answers, and unclear topics go to a human.
The live demo uses trusted cloud AI providers for quality and speed. A private self-hosted route can be planned for clients who need stronger data control, but it costs more to maintain.
MyClienta owns the widget, routing, prompts, guardrails, escalation flow, and monitoring. The client does not need to manage AI providers or technical updates.
When the agent is unsure, a request can be emailed to the right team with the visitor details, question, and selected package. That keeps support safe and traceable.
| Client question | Simple answer | Current setup | Production note |
|---|---|---|---|
| Is this your own product? | Yes. We provide the website voice widget, package setup, safety rules, escalation flow, and ongoing maintenance. | The client sees one product: VoicePilot. | The technical services stay behind the scenes and can be changed when needed. |
| Are conversations confidential? | Yes, we can configure the agent so conversations are handled as confidential support interactions. | Only the information needed to answer or create a support request is processed. | For production, we agree who can access messages, how long they are kept, and where requests are sent. |
| Is everything in-house? | No, not by default. The product is ours, but high-quality live voice needs specialist voice infrastructure behind the scenes. | That is normal today if the client wants good voice quality and fast replies. | A more private setup is possible later, but it costs more and needs extra monitoring. |
| Who maintains it? | MyClienta maintains the setup, updates, voice package choice, safety rules, and support routing. | The client does not need to manage AI tools or technical accounts. | Maintenance should be included in the monthly platform fee. |
| Can it avoid medical advice? | Yes. The assistant can be instructed to answer product/support questions and escalate clinical, safety, or unclear questions. | The demo is already written around this support-only rule. | For a real client, the final safety wording and escalation policy should be approved in writing. |
| Where does the voice processing happen? | The live demo uses secure specialist services behind the scenes so the voice sounds better and responds faster. | The client does not need to know or manage those services. | The final package is chosen based on cost, voice quality, and privacy requirements. |
| Can we keep data in Europe? | We can design the production setup around EU-friendly hosting, privacy rules, and short data retention. | This demo is a sales prototype, not a signed GDPR production configuration. | For German medical-device clients, final setup needs GDPR/DPA review and clear retention settings. |
| What is stored? | The demo keeps lightweight usage counters and sends package requests by email. Production can store transcripts and tickets only if the client wants reporting. | Audio does not need to be stored unless the client specifically wants recordings. | Durable storage should be added before selling a full SaaS dashboard. |
| What happens if the AI is unsure? | It should collect the question and contact details, then send it to a human instead of guessing. | The request email flow is already live and verified. | Real support escalation should include client-specific recipients and ticket storage. |
| Can it be connected to phone? | Yes, the same brain can later be connected to phone/SIP, but phone is not live on this demo yet. | Current demo is website voice/chat only. | Phone adds telecom cost, reliability work, and compliance risk. |
| Can it be fully private? | Potentially, yes, as a custom private package. But it should be sold only after testing the quality and reliability. | H Private is shown as planned, not live. | Only offer it after quality, latency, monitoring, and support process are tested. |
| What does the monthly fee cover? | Hosting, monitoring, configuration, prompt/knowledge updates, provider routing, basic maintenance, and the selected included-minute bundle. | The client chooses a flat bundle first: 500, 1,000, or 2,000 included minutes. | For 2,000+ minutes, quote custom pricing so usage stays profitable and controlled. |
After the client chooses a package, we install one script on their website and configure their knowledge base, escalation email, and allowed domains.
Starter Voice
This sends the request directly to MyClienta. We will review the selected package and reply with the setup steps.
Live browser voice call
Press start, allow microphone access, and ask a support question in English or German. The AI answers out loud with the selected voice route, then listens again.
Microphone permission is needed when you start.
Ready for a live browser voice call.
Enter PIN to see internal costs, margin, and demo usage.
Wrong PIN.
Public pricing is now flat bundles. The client sees 500, 1,000, and 2,000 included minutes, plus a 2,000+ quote option.
Formula: client monthly price = included minutes x (our real cost/min + EUR 0.10 margin/min).
Extra minutes: the public overage price is the same client extra-minute price shown in this table.
Profit rule: each used minute should keep at least EUR 0.10 margin before setup, support, tax, and refunds.
Shared platform cost is separate. For a real production dashboard with manuals/RAG/transcripts/usage, budget about USD 25/month for one shared Supabase project.
That project can serve the full route palette and multiple clients until traffic, data isolation, or compliance says we should split it.
Active today: Cloudflare Pages/Worker, browser speech APIs, Groq LLM, OpenAI Whisper/TTS, OpenAI Standard, Business Nova, OpenAI HD, ElevenLabs premium, and Deepgram Aura through the lab Worker.
Not connected here yet: Azure TTS, Cartesia, LiveKit, Telnyx, Supabase RAG, durable billing, manuals/PDF upload, real phone numbers, and the in-house Hetzner route.
Local/self-hosted: none of the AI models today. Only the product code and UI are ours right now. The in-house route is a future cost-control option, not a sellable live package yet.
No usage yet.