Your AI agent calls.
Objects. Books. Follows up.

Not for the demo. The browser voice chat and the simulated funnel run entirely on our infrastructure. When you move to production campaigns calling your real leads, you'll connect your own Twilio account — that's the only step that requires it, and we walk you through it inside the app.

The AI scrapes your client's website and learns their services, pricing, tone of voice, and FAQs. It's been tested on restaurants, dental, roofing, coaching, SaaS, home services, medspas, salons, and more. If your leads can be reached by phone, the agent can talk to them. If there's a niche we haven't covered, you can import your own knowledge base (PDFs, website pages, training docs) and the agent will learn from that too.

The agent discloses it's an AI at the start of every call — we tested this explicitly in the "Are You a Robot?" stress-test scenario above, and leads who are asked directly still convert because the honesty disarms them. If a lead says "stop calling," the agent apologises and hangs up immediately. The number goes on a do-not-contact list that propagates across every channel. No retries, no drip, no further outreach — ever.

AI disclosure is baked into every call by default, which is the current TCPA and state-level requirement for AI-initiated outreach. Compliance trails are created automatically for every opt-out, recorded, and retained for 7 years — exportable if you're ever audited. That said: you're still responsible for your own lead lists being consent-based. The tool handles the technical compliance; you handle the opt-in side.

The agent has hard guardrails against making financial guarantees, bashing competitors, or discussing topics outside its training. We specifically stress-tested against manipulation attempts — the "Trick AI Into Bad Statement" recording above is a real example of a caller trying to get the AI to call competitors scams, and the agent deflected all three attempts. Every call is transcribed and sentiment-scored so you can review anything flagged.

Short answer: US law. The 2024 FCC ruling made AI-generated outbound calls without prior express written consent per-se illegal under TCPA — statutory damages run $500–$1,500 per call. If a visitor entered someone else's number here (an ex, a journalist, a regulator), a single call could generate a real lawsuit. There's no way for us to verify consent from an anonymous web form, so the safe move is no real outbound calls from the demo — period. Browser voice chat is the honest alternative: you hear the exact same agent, talking in the same voice, handling the same objections, live. When you're inside the app with your own Twilio account and a lead list you know has opted in, that's the moment real outbound calls happen — with a compliance trail attached to every one.

The platform cost is fixed — flat pricing, no per-call charges from us. Your variable cost is whatever your own Twilio account charges for the call itself, which is typically 1-2 cents per minute for US calls. A typical AI call runs 60-90 seconds, so you're looking at roughly 2-3 cents in carrier cost per lead reached. Compare to paying a human appointment setter at $15-25/hour.

Under ten minutes if you already have a client URL and a list of leads. The "URL importer" drafts the funnel in 2 minutes. Reviewing the AI's script takes another 3-5. Connecting Twilio the first time is a 5-minute walkthrough. After that, launching additional funnels for new clients is roughly a 2-minute job each.

Live chat inside the app during business hours. A private community with 850+ agency owners running funnels through the platform. Wednesday office hours on Zoom every week. And email support within 24 hours, usually much faster. We've run 12+ JVZoo launches ourselves — we know what support looks like when you're trying to move fast.