Trust
What our AI will never do.
These aren't marketing promises -- they're hard rules built into the same agent that answers every call, the same one you can call yourself below.
Diagnose or promise a fix
The agent collects what a caller describes and gets it to your team. It's instructed never to diagnose the problem, guess at a cause, or tell a caller what's wrong with their equipment.
Quote a repair, install, or part price
Only your configured diagnostic or visit fee is ever stated as a real number. Every other price question gets routed to your technician for an honest quote after they've actually seen the job.
Ask for sensitive personal or financial information
No home-service call needs a Social Security number, card number, bank account, or ID number. If a caller starts to volunteer one, the agent stops them before they finish and never writes the digits anywhere.
Hide that it's an AI, or that the call may be recorded
Every single call opens with the agent identifying itself as AI and disclosing recording, with no exceptions -- not even for a frantic or emergency-sounding caller.
Refuse or stall a request for a human
The moment a caller asks for a real person, the agent stops what it's doing and starts the transfer. It doesn't argue, doesn't try to keep them talking, and doesn't claim a human is available if the transfer fails.
Text a caller without real, explicit consent
Consent has to be an unambiguous yes, asked as its own step, never a condition of getting service. Silence, a vague "mhm," or a caller talking past the question never counts as a yes -- and "don't text me" ends it immediately.
Tell a caller they're booked when they're not
The agent only says a time is confirmed when the scheduling tool actually returns a booked status. Otherwise it says dispatch will confirm timing -- it never claims a slot is locked in to make a caller feel better.
Turn a caller away on its own judgment
The agent isn't a coverage-verification system and won't flatly tell a caller you don't service their address, or send them to a competitor. If it's genuinely out of area, that's a call for your team to make -- not the AI's, mid-call.
Every rule above is enforced in the same instructions the agent runs on every real call -- not a separate policy page written after the fact. If a caller ever pushes back on a recording, the agent stops normal intake and offers a callback instead. If a caller says stop to texting, it stops, no second chances needed. We’d rather you hear this from the agent directly than take our word for it.
Hear it for yourself.
Call the live demo line or book time and we'll walk you through it.