The team behind your voice layer
The voice platform
we build with you
A documented, EU-hosted voice-orchestration platform, plus the engineers who own the layer and set your voice system up alongside your team. You build on a dashboard and API you can take over yourself. No lock-in.
const call = await voicedock.calls.create({ assistant: "support-nl", to: "+31 20 555 0123", from: trunk.eu, })
Proof, not badges
Real products already run on this stack
We don't ask you to be the first. Live voice systems run on VoiceDock today: our own product, an external partner company, and a white-label client. The same orchestration layer powers all of them.
Our own voice product
Built and run on VoiceDock end to end. We use what we sell.
An external partner company
A paying external partner that builds and runs its own products on our orchestration layer.
A white-label client
Runs voice AI under its own brand on infrastructure we maintain.
We build the orchestration layer that competes with Vapi
VoiceDock isn't a wrapper around someone else's platform. We are the developers and maintainers of the orchestration layer itself. When a call fails, we don't file a ticket with a vendor, we open the SIP-trunk logs and debug at the deepest level. A Vapi-reselling agency can't do that, because they don't own the stack.
We read the SIP-trunk logs
This is the level a wrapper agency can't reach. We can, because we built the trunk integration.
$ voicedock logs --call 4f9c… --debug
[sip] INVITE 200 OK · trunk=telnyx-eu-1
[sip] re-INVITE · packet loss 4.2% detected
[rtp] codec=OPUS · jitter 38ms · climbing
[orchestrator] barge-in dropped · buffering 320ms
[fix] jitter buffer adaptive → call recovered
[rtp] jitter 6ms · stable
✓ resolved at the SIP layer — not a black boxWe design it together, you own it when it's live
Three steps, with our engineers next to your team. Not a self-service form and a "good luck".
We map your call flow together
In a working session we design the agent, the tool calls, and the provider stack around your use case, language, and compliance needs. You watch how every decision is made.
We build and connect it with you
We wire up phone numbers, SIP trunks, and webhooks, and go live together. When something breaks in telephony, we debug it at the log level. That's our layer.
You run it, we stay on the stack
You manage everything from the dashboard and API: assistants, calls, transcripts, analytics. We keep maintaining the layer underneath, so you scale without inheriting an ops burden you didn't sign up for.
A platform you operate, not a black box
See and steer every call
Live transcriptions, tool-call execution, and real-time intervention from one dashboard. The same orchestration layer we build and maintain ourselves, so it does things a reseller platform can't.
Learn about orchestrationYour keys, your models
Use our managed infrastructure or bring your own API keys. Switch between Deepgram, ElevenLabs, OpenAI, Mistral, or any supported provider without touching your code.
Learn about BYOKA dashboard and API you own
Everything we set up with you is reachable through a clean REST API and dashboard. Manage it yourself, hand it to your team, or keep us on retainer. The choice stays with you.
Explore the APIconst response = await fetch(
'https://api.hmsovereign.com/api/v1/assistants',
{
method: 'POST',
headers: {
'Authorization': 'Bearer org_api_key_...',
'Content-Type': 'application/json'
},
body: JSON.stringify({
name: 'Support Assistant',
system_prompt: 'You are a helpful...'
})
}
);
const assistant = await response.json();Built for developers who ship fast
A clean TypeScript SDK and RESTful API. Create an assistant, start a call — that's it. No abstractions to fight.
import { HMSSovereign } from "hmsovereign"
const client = new HMSSovereign({ apiKey: process.env.HMS_API_KEY })
const assistant = await client.assistants.create({
name: "Customer Support",
model: {
provider: "openai",
model: "gpt-4o-mini",
systemPrompt: "You are a helpful support agent..."
},
voice: {
provider: "elevenlabs",
voiceId: "rachel"
},
transcriber: {
provider: "deepgram",
model: "nova-3",
language: "nl"
}
})
console.log(assistant.id)npm install hmsovereignv1.0.1Built for any conversation you need
From inbound support to outbound campaigns, VoiceDock powers voice agents across industries, all on the same orchestration core we build and maintain ourselves.
Customer support
24/7 inbound agents that escalate seamlessly to humans. Order lookups, refund flows, and FAQ handling.
Appointment booking
Reminders, confirmations, and reschedules. Wire it into your CRM and calendar of choice via custom tool calls.
Outbound campaigns
Lead qualification, surveys, follow-ups. Concurrent dialing with intelligent abandonment.
Hotels & restaurants
Reservation handling and front-desk overflow. Mews PMS integration in production for early customers.
Sensitive workflows
Insurance intake, internal helpdesks, and regulated industries. On-premise deployment available for full data control.
Internal tools
IT helpdesks, HR FAQs, sales enablement. Hook into your existing tools via custom function calls.
Every provider, one API
Mix and match STT, LLM, and TTS providers per agent. Switch providers with a single config change, no code changes needed.
Bring your own keys. Use your existing API keys and pay providers directly. We charge only for orchestration, €0,07/min, with no markup on provider costs.
Not a tool. Not an agency. Both
The giants hand you an API and walk away. A wrapper agency builds it for you but can't go deep when telephony breaks. We build it with you and stay on the layer underneath.
Self-serve platform
The big dev platforms
- Scales
- You're on your own
- US-hosted
Specialist agency
A reselling wrapper shop
- Builds it for you
- Stuck at the surface when telephony breaks
- You don't own the result
VoiceDock
the third way
- Builds it with you
- Owns the stack, debugs at the SIP level
- EU-hosted, you own the dashboard and API
How we work together
Intake call
We map your call flow and the right provider stack.
We build and wire it with your team
Live deployment, with telephony debugged at the log level.
You operate it, we stay on call
You run it on infrastructure you own. We keep maintaining the layer.
Built for teams ready to invest in a voice system that actually ships, not a weekend prototype.
Your platform layer, hosted in Europe.
Our orchestrator, dashboard, and storage run on EU infrastructure. Pair it with EU-resident model providers for an end-to-end European stack — or pick whatever fits your use case. We help you choose.
Managed Cloud
Deploy in minutes on our EU-hosted infrastructure. No US cloud providers in the path.
- Zero ops overhead
- Auto-scaling
- EU data residency
Private Cloud
Dedicated instances in the EU cloud provider of your choice. Full isolation in your own VPC.
- EU cloud providers
- Your compliance
- Custom SLAs
On-Premise
Full black-box installation on your hardware. Air-gapped networks supported.
- Your hardware
- Air-gapped capable
- Zero external calls
Deep tech for the most demanding environments
Banks, healthcare providers, government agencies -- we serve organizations where compliance isn't optional. HMS-Sovereign v1 runs entirely within your infrastructure with zero external dependencies.
EU residency for the data we control
Audio, transcripts, recordings, and metadata that flow through our orchestrator stay on EU infrastructure. Zero US cloud providers in our platform layer. Model-provider choice is yours and we'll advise on the most compliant stack.
Build your own voice product on our stack with us behind it
For agencies and software companies who want to sell voice AI under their own brand. You get the branded portal, billing, and usage tracking, and the team that built the orchestration layer maintaining it underneath. We grow through partners, not self-service volume.
Fully branded client portal
Custom domain, logo, and colors. Your brand, our infrastructure.
Integrated Stripe payments
Automatic billing with configurable margins. Set your pricing and go live.
Real-time usage tracking
Client self-service portal with call history and credit monitoring.
Flexible configuration
Client-managed or agency-managed API keys. You choose the model.
White-label configuration
Full control over branding, pricing, and payment processing
Pay for the platform. Invest in the build.
The platform is €0,07/min, keys-included or BYOK. The value is the team that builds it with you.
We design, build, and ship your first voice system with your team, on a platform you take over afterwards. For organizations that want it done right, not done alone.
Plan a callScoped per engagement — a multi-week build for teams serious about voice.
The self-serve platform. Bring your own keys for STT, LLM, and TTS, and orchestrate on our layer.
+ your own provider costs
*Fair use policy applies
For scale-ups and enterprises with significant call volume. Volume-based pricing available.
Flexible pricing for your needs
Competitive volume-based pricing. Let's discuss what works best for your call volume.
Full sovereignty. On-premise, air-gapped, or private cloud deployment.
Full deployment support, training, and ongoing consultancy. Source code escrow available.
Enterprise services
Everything you need for mission-critical voice AI
Self-Hosted
On-Premise
Private Models
Custom Integrations
Training & Consultancy
Dedicated Support
Source Code Escrow
Custom SLA
Need something specific? We've built voice AI for banks, healthcare providers, and government agencies.
Let's discuss your requirementsBring us your hardest voice use case
Book a call and we'll map your call flow, the right provider stack, and what a first deployment looks like. You leave with a concrete plan, whether or not you build it with us. No funnel, no obligation.