Frequently Asked
Questions.

No. We strictly bypass traditional touch-tone IVR bots. We deploy neural streaming models with conversational breathing pauses, dynamic pitch inflections, and natural filler sounds. Over 95% of callers speak normally and never realize they are speaking with an AI.

Our full round-trip conversational latency is engineered to stay strictly under 500 milliseconds. By combining Deepgram Nova-2 streaming speech-to-text with optimized LLM token streaming, the conversation flows at normal human conversational speed without awkward pauses.

Callers can interrupt at any moment. Our agents feature bi-directional full-duplex audio. The instant a caller begins speaking or objects, the agent immediately stops audio playback, processes the user's interruption, and responds in context just like a real person would.

We apply strict deterministic guardrails. The AI is bounded strictly to your verified knowledge documents, pricing lists, and operational FAQs. If a caller asks something out of scope, the AI is programmed to state: "Let me have one of our specialists confirm that for you," and either initiates a warm transfer or creates an urgent follow-up task in your CRM.

No. You keep your existing numbers and phone provider. We configure conditional call forwarding (e.g., forward when your front desk is busy, when a call goes unanswered after 2 rings, or after 6:00 PM). It integrates into your existing business line with zero downtime.

We connect directly via secure APIs and webhooks to Google Calendar, Calendly, GoHighLevel, HubSpot, Salesforce, ServiceTitan, Housecall Pro, Dentrix, Reapit, and custom webhook endpoints. It performs two-way slot checking in real-time while on the phone with the customer.

Yes. The AI can trigger immediate outbound SMS or WhatsApp messages while the caller is still on the line (e.g., "I’ve just sent the booking confirmation and address straight to your mobile number").

The agent performs an instant live warm transfer. It dials your on-call technician or receptionist, whispers a brief context summary of who is calling and why, and bridges the caller without them having to repeat their situation.

The system can effortlessly handle 200+ simultaneous phone calls without busy tones, hold music, or dropped calls. If your business runs TV ads or experiences a morning surge, every caller gets answered on the first ring.

A standard single-pipeline deployment takes 10 to 14 business days. This includes auditing past call logs, crafting custom prompt guardrails, integrating your CRM APIs, acoustic voice tuning, and rigorous edge-case stress testing before going live.

A full-time in-house receptionist typically costs £25k–£35k+ per year (plus taxes, benefits, and sick days) while only covering 8 hours a day. Outsourced call centers charge high retainers and deliver poor conversion rates. An AI voice agent delivers true 24/7/365 coverage for a fraction of that cost, operating at pennies per conversation minute.

Yes. All audio data streams in-flight with TLS encryption, and stored recordings/transcripts are encrypted with AES-256. We configure call consent disclosures (e.g., "This call is recorded for quality purposes") and adhere to GDPR data retention policies.