The system

One private system underneath every site you run.

Not another binder. Not another app. A connected system that keeps every site aligned to your policies, safe to use with sensitive information, and protective of the people doing the work. Here's what it's made of.

What we do

Six parts, built to work as one.

Each part solves a specific problem you already live with. For each one: what it is, the problem it solves, and what it looks like on a real shift.

1

A private, confidential AI network

What it is. Your own AI, walled off from the public models. Your staff get the help they'd otherwise go looking for, but inside your own walls — with data kept in Canada and the whole thing built around BC privacy law.

The problem it solves. Right now your team is reaching for ChatGPT and the rest. The moment someone pastes a case note in to "clean it up," personal information has left your control. This gives them a safe place to do the same work.

On shift. A worker drafts a report, summarizes a hard day, or asks how to phrase something difficult — and never has to choose between getting help and protecting confidentiality.

Solves: staff pasting case notes into ChatGPT

2

An operational dashboard

What it is. One clear picture of where every site stands — incidents, reporting, training, compliance — in a view that's ready for you and for the standards, funders, regulators, or accreditation bodies you answer to.

The problem it solves. Today you find out where things really stand at the audit or the incident review — not before. There's no single place that shows you the real state of your operation across sites.

On shift. You open one screen and see which site is behind on reporting, where an incident needs follow-up, and who still needs training — instead of chasing it by phone and email.

Solves: flying blind, finding out at the audit

3

An employee agent for onboarding & training

What it is. An agent that onboards and trains every new hire the same way — your way — and keeps doing it consistently even when you're short-staffed and leadership is stretched.

The problem it solves. When there's no time, new staff learn by osmosis. Two people in the same role end up working two different ways, because training depends on who happened to be on shift that week.

On shift. A new worker is walked through your onboarding, your expectations, and your way of doing the work — on day one, regardless of how busy the team is.

Solves: onboarding by osmosis, inconsistent training

4

A site-by-site agent aligned to your standards

What it is. An agent trained on your own policies and procedures and aligned to the standards you're held to, so staff can get the right answer in the moment — at the site, on shift, when it actually matters.

The problem it solves. The binder sits on a shelf and no one opens it mid-shift. So policy gets followed when it's remembered and skipped when it isn't, and you only learn the difference after something goes sideways.

On shift. A worker asks what the procedure is for a specific situation and gets your answer — the one that matches your policy and external obligations — instead of guessing.

Solves: the binder no one opens, staff not following policy

5

Unified notes & reporting

What it is. One common methodology for how your people collect, summarize, and report — aligned to your organization and to what your funders, regulators, and accreditors expect to see.

The problem it solves. Every site reports differently. Notes from one program or unit don't line up with another, with your own standard, or with what reviewers expect to see. Reviewing it all takes time you don't have.

On shift. Staff record what happened in one consistent way, and it comes out the other end already in the shape your organization and accountability bodies need.

Solves: every site reporting differently

6

Psychological health & safety, throughout

What it is. The whole system is built on Canada's national standard for psychological health and safety in the workplace (CAN/CSA-Z1003), so it protects the people doing the work — not just the paperwork.

The problem it solves. Burnout, turnover, gossip, and vicarious trauma are the quiet cost of this work. They take the staff you can least afford to lose, and no single tool fixes that — it has to be built into how everything else works.

On shift. Load comes off your people, the work is more predictable, and the conditions your team works in are treated as something to protect — by design, not by poster.

Solves: burnout, turnover, gossip, vicarious trauma

Confidential by design

Why a private AI network — not a public one.

You hold sensitive personal information about people in care. That single fact decides how this has to be built. Here's why public tools are a real risk — and what makes the private network different.

Why public tools like ChatGPT are a real risk

When a staff member pastes personal information into a public AI tool, that information leaves your control. It can be stored outside Canada and used in ways you never agreed to. For care providers, that's not a small slip — it's a confidentiality breach.

This isn't hypothetical. BC's privacy regulator, the Office of the Information and Privacy Commissioner (OIPC), took part in a joint investigation that found OpenAI — the company behind ChatGPT — was not complying with Canadian privacy law, because it used people's personal information without proper consent. The OIPC has also published guidance warning organizations to be careful about putting personal information into these tools.

How the private network is different

We build it the opposite way around — privacy first. We call it Sovereign AI: your own model and the apps your team uses, running on hardware you control, with your data kept in Canada.

  • Designed to keep data in Canada. Information is designed to stay where the law expects it to be, not on a public model's servers, in line with BC's guidance on disclosures of personal information outside Canada.
  • Built around BC's privacy framework. It's designed around the Personal Information Protection Act (PIPA) — the law that governs private organizations like yours — and the Freedom of Information and Protection of Privacy Act (FOIPPA), which governs the public bodies many care providers report into.
  • Staff never paste into public tools. Because they have a safe place to get help, there's no reason for personal information to ever go into ChatGPT or anything like it again.

How your data actually stays private

This is the part most "private AI" claims skip over, so here's the honest version.

  • Day to day, your data never leaves. The model runs locally. Your information is never sent anywhere to answer a question.
  • It still gets smarter — safely. When we improve the model, we use anonymized examples only — never your real data. The outside systems involved in that process never see anything from your organization.
  • You hold the keys. The data, the model, and the keys stay with you — in Canada, outside the reach of US data-access laws like the CLOUD Act.

Two ways to run it

  • On-premise. A dedicated box on your own network. Your data physically never leaves the building.
  • Managed in Canada. We run your dedicated machine in a secure Canadian facility, so you manage no technology.

Either way the AI is yours — your model, your data, your dedicated hardware. Never shared with another organization.

What's included

  • A model tuned to your work — trained on your documents and procedures, not a generic chatbot.
  • Your own apps and assistants, in an environment only your people can reach.
  • Real privacy controls — full encryption, role-based access, and a complete audit trail.
  • Ongoing care — security updates, retraining, and support — on a subscription.

Pricing is a one-time build and setup, then a monthly subscription for security, updates, improvement, and support. What it comes to depends on your size and how you deploy — so we quote it after a conversation.

Book your free, confidential 45-minute conversation

The point of all this isn't fear — it's two things you already care about: protecting the confidentiality of the people you serve, and protecting your accreditation or regulatory standing, which depends on handling their information the right way. Modern AI is too useful to ignore. This is how you get it without the compromise.

Built in, not bolted on

How psychological health & safety is built in.

The whole system is built on CAN/CSA-Z1003, Canada's national standard for psychological health and safety. Protecting your people isn't a separate program you have to remember to run — it's the foundation the other five parts sit on.

Sources

Where the claims on this page come from.

OIPC finding that OpenAI / ChatGPT did not comply with Canadian privacy law — BC OIPC, Joint Investigation of OpenAI (2026); OIPC guidance on privacy and AI — BC OIPC, Privacy Guidance on Generative AI; Personal Information Protection Act (PIPA) — BC Laws, PIPA; Freedom of Information and Protection of Privacy Act (FOIPPA) — BC Laws, FOIPPA; data residency / disclosures outside Canada — Province of BC, Disclosures Outside Canada; US data-access law (the CLOUD Act) — US Department of Justice, CLOUD Act Resources; psychological health & safety standard — Government of Canada, psychological health and safety at work / CSA Z1003; BC accreditation policy for social-service providers — Province of BC, Accreditation Policy.

Start here

It starts with a conversation.

No pitch, no cost — 45 minutes to get an honest look at where your organization is and what would actually help. You'll leave with something useful whether or not we work together.

Book your free, confidential 45-minute conversation

Confidential. 45 minutes. Virtual.