Launch your own voice AI product in days, not months. Your branding, your pricing, your clients. $0.03/min platform fee. Bring your own keys. No bulk minimums. No per-seat upcharges.
Most “white-label” voice AI products are single-tenant tools with a logo upload field. Burki was built for agencies from the ground up.
Each of your clients gets their own organization, isolated data, scoped API keys, and per-tenant rate limits. Built into the platform from day one — not bolted on.
Custom subdomain, your logo and colors in the dashboard, your domain on outbound caller ID and webhook URLs. End-clients never see Burki.
$0.03/min platform fee + actual provider pass-through (you can see every line item). Set your own client pricing on top — typical agency markup is 3–5x.
Bring your own Twilio/Deepgram/ElevenLabs/OpenAI keys to bypass markup entirely, or let Burki manage providers and bill through. Per-tenant override supported.
Native support for marking up provider costs per client, applying volume discounts, white-labeled invoices, and Stripe Connect for direct client billing.
Every dashboard action is also an API call. Embed Burki inside your own SaaS, automate tenant provisioning, and build custom client portals on top.
Most white-label voice platforms lock you into one provider per layer and pocket the spread. Burki gives you the whole stack.
| Layer | Burki | Typical white-label |
|---|---|---|
| Telephony | Twilio, Telnyx, Vonage — pick or mix per tenant | Locked to vendor's chosen carrier |
| STT | Deepgram, Azure (15+ supported) | 1–2 fixed providers |
| LLM | OpenAI, Anthropic, Google, Groq, OpenRouter (50+) | OpenAI only or limited list |
| TTS | ElevenLabs, Cartesia, Azure, Deepgram, Kokoro, Resemble (15+) | ElevenLabs only, marked up |
| Memory + RAG | Built-in pgvector, episodic + semantic + procedural | Often missing or extra-cost add-on |
| IVR Mapping | Free IVR Explorer included | Not available |
| Per-tenant branding | Subdomain, colors, logo, caller ID | Limited or paid add-on |
A typical agency client running 10,000 minutes/month.
Charge what your market bears. We don't cap or police client pricing.
It means your agency operates Burki as your own product. Your clients log into your subdomain, see your logo and colors, get calls from your phone numbers, and receive invoices from your billing entity. Burki never appears in their world.
$0.03/min platform fee on top of actual provider costs. There are no per-tenant licensing fees, no minimums, and no markup on provider passthrough. You set whatever client price you want — typical agencies charge $0.15–$0.40/min and capture the spread. With BYO keys you can bring provider costs down further.
Yes. Per-tenant key configuration is built in. A client can BYO some providers and let your agency manage others. This is how many agencies handle larger enterprise clients who already have provider relationships.
Vapi and Retell were built single-tenant first. White-label features were added later and are limited (no real per-tenant theming, no custom domain on dashboard, capped provider choice). Burki was designed multi-tenant from the start — every model is org-scoped, every API call is tenant-aware, and the dashboard branding is fully configurable.
Yes — we're onboarding our first 25 agencies under our Founding Agency Program with locked-in pricing, white-glove onboarding, and direct Slack channel to the engineering team. Apply at /agencies.
Most common: AI receptionists for healthcare practices, real estate lead qualifiers, restaurant order-takers, automotive service schedulers, insurance claims intake, and outbound cold-calling agents. Agencies typically deploy 5–50 voice agents per client across multiple verticals.
First production agent: same day. Full white-label setup with custom subdomain and branding: 1–2 days. Ten clients onboarded: under two weeks for most agencies. Compare to building from Twilio + OpenAI directly, which takes 4–8 weeks before you have something resellable.
Apply to the Founding Agency Program for locked-in pricing, white-glove onboarding, and direct engineering Slack access.