A Canadian AI readiness assessment that examines your permissions and data, not a questionnaire.
One Canadian team examines what your AI would actually reach, so the AI capability assessment rests on what's configured in the tenant rather than on what a questionnaire remembers.
We scan what's overshared, stale, and unlabelled across the tenant and the data estate, because your AI inherits every one of those. Data readiness for AI is the dimension that decides whether a pilot is safe to start.
Licensing position, technical prerequisites, and the sharing and labelling cleanup that has to happen before a Microsoft 365 rollout. Copilot readiness is the most common reason Canadian firms call us, and it's a permissions job before it's a licensing one.
Infrastructure, skills, and governance scored against evidence, with each gap priced. The AI gap analysis says what to fix before a pilot, what can wait, and what's already fine as it stands.
40% of organisations delayed a Copilot rollout by three months or more because of oversharing (industry surveys and Gartner, 2026). Finding it before the licences land is the cheap version of that discovery.
IN PRODUCTION
We'd done the free vendor questionnaire and scored ourselves 'mostly ready'. SMEnode plugged into the tenant and found 40,000 files shared with everyone in the company, including payroll. Copilot would have surfaced all of it on day one. The score wasn't wrong about our strategy. It just couldn't see our permissions.
Free self-assessments score what you believe. This one reads what's configured, and it's written by the engineer who runs the scan.
Every AI capability assessment covers the same dimensions: data, infrastructure, skills, governance, strategy, culture. The dimensions aren't the hard part. The hard part is that a questionnaire asks your team whether data access is well controlled, and the answer is always a considered yes. Then you plug in an assistant and find the finance folder was shared with the whole company in 2021. AI doesn't grant new access, it just makes existing access findable. That's why an honest readiness score needs someone reading permissions, not collecting opinions.
The risk sits in three places, and they're specific. Content shared with everyone in the organisation through a default nobody changed. Broken permission inheritance, where a locked site holds an unlocked folder. And stale documents that are wrong but confident, which an assistant will quote back as fact. Roughly 16% of business-critical data is overshared in an average Microsoft 365 tenant, and 40% of organisations delayed a Copilot rollout by three months or more because of it (source: industry surveys and Gartner, 2026).
19.2% of Canadian businesses used AI in production in Q2 2026, up from 12.2% a year before, and 13.4% name cybersecurity or privacy as the barrier that limits them (source: Statistics Canada, 2026-05-27). Those firms are right to worry, and a readiness assessment is the cheap way to find out whether the worry is justified. PIPEDA applies to everything the model can reach, and Quebec's Law 25 applies if you hold data on customers there.
Four steps, and none of them is a survey. We ask for read-only access to the tenant and the data estate, then go and look. Clients sometimes push back on that, because a free tool takes twenty minutes and reading a real tenant takes proper time. The difference is that only one of them can tell you your payroll folder is world-readable. We score six dimensions, and the findings that change decisions are almost always in data and permissions, so that's where most of the effort goes.
Read-only access to the tenant, the sharing reports, and the data stores in scope. Nothing changes, nothing moves, and you can watch every query we run.
Oversharing, broken permission inheritance, stale content, and unlabelled sensitive data. This is the list a questionnaire structurally could not produce, because it's read out of the configuration rather than out of memory.
Data, permissions, infrastructure, skills, governance, and use case fit, with evidence attached to each score. A score with no artefact behind it is an opinion, so every one of ours points at something we read.
The fixes that have to land before any pilot, priced and ordered, plus what's fine exactly as it stands. That second list is usually longer than the client expects, and it's the part that makes the budget conversation easy.
Answers first, including when the free tool is enough. An architect takes the call.
LET'S CONNECT
A senior engineer replies within an hour, 24/7.