FAQ
Questions Canadian Businesses Ask About AI Agents
Straight answers about what AI agents do, where your data goes, what happens when the model is wrong, what it costs and who owns the result. If your question isn't here, ask it on a free 30-minute audit call.
AI agents: the basics
What is an AI agent, in plain terms?
An AI agent is software that reads a piece of incoming work, decides what needs to happen, and does it using your existing tools. It might read an emailed purchase order and create a draft sales order, or answer a phone call and book an appointment. It works within the permissions you give it and asks a person to approve anything that matters.
How is an AI agent different from a chatbot or a Zapier workflow?
A chatbot answers questions but rarely does anything in your systems. A Zapier or Make.com workflow follows fixed rules and breaks when the input doesn't match what it expects. An agent can read messy input, like a scanned PDF or a rambling email, decide what to do, and act in your tools. Most of our agents combine both: a workflow for the fixed steps, and an AI model for the reading and deciding.
Which AI models do you use?
We pick per job. For hosted work we mostly use Claude from Anthropic and models from OpenAI. For private work we run open-weight models such as Llama, Qwen and Mistral through Ollama. Voice agents run on Retell AI or Vapi. We test on your real examples before choosing, and we build the agent so the model can be swapped later.
Will AI agents replace people on our team?
We don't scope projects that way. We pick the repetitive work your team does reluctantly, like re-keying orders, sorting a shared inbox or chasing quotes, so the same people spend more of their day on customers and decisions. A person still approves anything that matters.
Mistakes, approvals and logs
What happens when the AI model gets something wrong?
We plan for it. Agents check their work against rules, such as totals adding up or a customer existing in your system, and stop to ask a person when they aren't confident. Actions that matter wait for approval, so most mistakes are caught as drafts. The activity log shows what went wrong, and we keep a test set of your real examples to re-run after every fix.
Do agents need a person to approve their work?
For anything that sends money, signs something or emails a customer, yes by default. You can change that per action. We recommend starting with approvals on almost everything, then letting routine actions run on their own once the log shows the agent gets them right.
What do you log, and who can see it?
Every run records what started it, what the agent read, what it decided and why, each tool it used, what changed, and who approved it. The log sits in a place your team can open without us, such as a SharePoint list, a Google Sheet or a database table. You decide who has access and how long entries are kept.
Data, privacy and hosting in Canada
Is our data safe? What about PIPEDA and Quebec's Law 25?
We design with PIPEDA and Quebec's Law 25 in mind. Each agent gets its own account with only the permissions its job needs, sends the model only the data the task requires, and logs every action. Before you sign, we document where each piece of data goes, so your privacy officer or lawyer can review it. We don't claim any certification.
Can the agent and its data stay in Canada?
Yes. We can run the model and the agent on Azure Canada Central, AWS ca-central-1 or your own hardware, using open-weight models such as Llama, Qwen or Mistral. Hosted models like Claude and OpenAI may process data outside Canada, so for sensitive data we recommend a private agent.
Will our data be used to train AI models?
We use the business APIs from Anthropic and OpenAI, not their consumer chat apps. Under their business terms, both say they don't train their models on API data by default. If you'd rather not rely on a provider's terms, a private agent keeps the model on servers you control.
Working with us, cost and ownership
Do you really build a prototype for free?
Yes. I'll build you a working prototype in 5 business days, free, with no obligation. It starts after the free 30-minute audit, once we agree on the one job to prototype. It's one workflow or one agent, built against your real tools in a test setup or on sample data you provide, and it doesn't touch live customers or production data. It's working within 5 business days of the day we have the sample data and test access, and you see it running on a video call. If you don't go ahead, you owe nothing and keep a written summary of what it does and what the full build would cost. The working prototype itself stays with us. If you go ahead, we build it into production at the fixed price on our pricing page, and the prototype isn't charged on top. We take on 4 free prototypes a month, first come, first served.
How long does it take to get an agent live?
You see a free working prototype within 5 business days of us having your sample data and test access. Once you approve the fixed price, the production agent is typically live in about 2 weeks, including a staging version you test before go-live. Private agents that need new hardware take longer, and the scope says so up front.
What do you need from our team?
One person who knows the job, a few dozen recent examples of the work, your unwritten rules, an account for the agent in each system, and a couple of hours to test the staging version. Nobody needs to be technical.
Do you work with businesses outside Ontario?
Yes. We're based in Ontario and work with businesses across Canada. Calls are by phone or video.
How much does an AI agent cost?
The audit and the working prototype are free. After that, every project gets one fixed price in writing before work starts. Starting prices for a first agent, a single automation and monthly support are on our pricing page. Model usage, voice minutes and software subscriptions are separate running costs, and we estimate them from your volume in the scope.
Who owns the agent you build?
You do. The workflows, prompts, instructions and documentation we create for your agent belong to your business. Wherever we can, the agent runs in accounts you own, such as your Microsoft 365 tenant, your n8n instance, your model provider account or your cloud subscription, so you aren't locked in to us.
What happens if we stop paying for monthly support?
Monthly support is month to month, and you can cancel anytime. The agent keeps running. Because it lives in your accounts, cancelling monthly support stops our monitoring and updates, not the agent. You keep the documentation, and your team or another developer can maintain it. If we host any part of it for you, we help you move it to your own account first.
AI made building about 5x faster. Get the savings.
Book a free 30-minute call. If there is a good first build, I will make you a working prototype in 5 business days, free, with no obligation.
Get a Free Working Prototype