One structure you'll always understand.

Those route calls or take basic messages. We configure an agent that actually knows your business — answers real questions, books and qualifies, and takes scoped actions — then hands off to a person when that's the right call.

That's exactly why this is a managed service, not a raw tool. We ground the agent in knowledge you approve before launch, force it to escalate when it's unsure, limit its permissions, gate anything irreversible behind human approval, and monitor performance closely — heavily so for the first two weeks.

Compare it to the cost of missed calls, owner hours lost to repetitive work, or another hire. Pricing scales by how intensively you actually use the agent, so you only pay for the volume you have — and building it yourself, or piecing it together from raw tools, usually costs far more once you count the time and the gaps.

You can experiment. Most businesses then find they need proper telephony, permissions, monitoring, knowledge grounding, and someone accountable for the results. That accountable, configured, managed layer is the service.

Your agent keeps running — we don’t cut service off mid-month for going over. The extra is billed as overage at the rate in your agreement, shown as its own line on the invoice with the rate printed on it. You’ll have heard from us before that happens: we send a notice at 75% of your allowance with a projection of where the month is heading, again at 90%, and at 100%. If you’re consistently running heavier than we scoped, we’ll move you to the right profile instead of letting you pay overage every month — that’s usually cheaper for you.

Yes. You can add usage credit in the portal — $100, $250, $500, or $1,000, with bonus credit on the larger amounts. It draws down against usage beyond your allowance automatically, applies to your next invoice, and doesn’t expire. It’s optional; plenty of clients never use it.

You own all of it — documents, recordings, transcripts, and configuration. The platforms we build on — xAI and Anthropic — maintain SOC 2 Type 2 compliance, and your data is never used to train their models. We process it only to deliver your service, and one client's knowledge never touches another's.