Fair questions. Straight answers.

Because unscoped quotes are how AI projects end up in the 40% that get cancelled. Every engagement is priced from a written scope you approve first — one workflow, one number it has to move, fixed cost, fixed timeline. The audit and the scope are free, so you see the price before you commit a dollar.

Only the underlying tool subscriptions, which you pay directly — typically $100–300/month depending on volume. No markup, no ongoing fees to us unless you choose a support plan.

Every build is measured against the baseline we record in week one — response time, booked appointments, admin hours. You get the before/after report in writing. If the number didn’t move, that conversation happens with data on the table.

The smaller the team, the more a missed lead or a wasted hour costs proportionally. That’s why we start with one two-week workflow, not a platform. And if the audit says you’re not ready, we’ll tell you — the playbooks are free for exactly that case.

It’s constrained, not creative: it answers from your approved knowledge base, and anything outside it — clinical questions, legal advice, complaints — gets handed to a human immediately, with the full conversation attached. You review every answer it’s allowed to give before launch.

We build PIPEDA-conscious by default: minimal data collection, Canadian data residency where available, business-grade AI accounts that don’t train on your data, 2FA everywhere, and a written map of exactly what is stored where — handed to you at launch.

No. We only deploy tools on business terms where your data is excluded from model training, and the AI house rules we set with your team make sure staff don’t paste client details into consumer accounts.

The system is designed so the AI handles logistics (hours, booking, intake) and never gives professional advice. You get the written data map and conversation logs, which is more documentation than most practices have for their current voicemail.

Yes — every account, login, automation, and line of code is set up in your name from day one. If we part ways, everything keeps running and it all stays yours.

Almost never. AI Rescue starts with what you own: we inventory it, give each tool one scoped job with one number, and either make it pay or give you the case for cancelling in writing.

Yes — when the front door is the bottleneck. We build sites that book rather than collect compliments, client portals, and web applications wired into your intake and CRM from the first day rather than integrated afterwards. Two of our own products are public and still ours to maintain: ComplyWise, third-party risk management aligned to OSFI B-10 for Canadian banks, credit unions and payment providers, and CollabNorth, invoicing that resolves GST, HST, PST, QST and RST by place of supply. We also built Infinite Immigration, the practice site for a licensed immigration consultancy. Running a regulated product ourselves is why we are comfortable building one for you. Native iOS and Android is scoped case by case. And we will tell you when you don't need one: plenty of builds here have run entirely on the website and phone number a client already had, which is why they took two weeks instead of two months.

Almost certainly. We build around your calendar, your phone number, your practice software or CRM — rather than forcing a switch. The scope document lists every connection before you approve it.

The agent answers instantly, handles the routine 80%, and hands anything sensitive straight to your staff. Most callers just notice they finally got an answer at 9pm. And it introduces itself honestly — no pretending to be human.

Nothing dramatic. The system is in your accounts, so it keeps running. You lose our monitoring and optimization, not the machine.