Everyone is talking about AI.
You don't know where to start.
Yourcompetitorsaremovingahead.
The gap widens.
AI audit for businesses
To make a real difference, start with an audit.
An audit of your technology environment shows where AI creates value for you, and in what order to implement it.
Our method in five steps
Five steps, five deliverables: from the map of your systems to the funding plan.
- Step 01
Mapping the technology environment
A map of your technology environment: what you have, what talks to what, and where the information stops.
- Step 02
Identifying the bottlenecks
Your bottlenecks ranked by estimated annual cost and by difficulty of correction.
- Step 03
Where and how AI can fit
A list of qualified use cases, each tied to a measured bottleneck.
- Step 04
A precise action plan
A phase-by-phase action plan, executable without us.
- Step 05
A funding plan for the project
The funding plan for the project: which programs, for which phase, in which order.
- DiagnosticFull AI audit
- ScopingQualifying a use case
- DataData preparation
- ProductionFrom prototype to real use
- FundingBuilding the funding file
- DefenceDefending an AI-related claim
And the audit can be funded.
Grants and tax credits for AI
An artificial intelligence project is almost never funded by a single program. The fees for an audit or a diagnostic fall under a first set of grants, paid by Québec, by a city or by a sector cluster; the development that follows falls under tax credits — SR&ED federally, and Québec’s CDAE for its artificial intelligence stream; integration in an industrial setting and access to computing power fall under separate programs again. The five programs below are the ones that most often fund the audit stage itself. Their stacking rules and calendars differ, so the filing order determines the amount actually obtained.
- 01Investissement QuébecESSOR — sub-stream 1CUp to $50,00050% of the fees for a study or diagnostic carried out by a third party
- 02PROMPTTechnology requirements documentUp to $5,000PROMPT’s maximum contribution — 50% of the fees set out in the pre-approved service offer
- 03Ville de QuébecSupport vouchersUp to $30,000Maximum non-repayable contribution per company per year, for up to 30% of eligible expenses
- 04Canada Economic DevelopmentIRRT — Regional Tariff Response InitiativeUp to $1MMaximum contribution per project
- 05Government of AlbertaAlberta Manufacturing Productivity Grant (AMPG)Up to $30,00050% of eligible expenses
These amounts are ceilings, not entitlements: each program has its own criteria, filing dates and stacking rules, and the same expense can never be funded twice. Several of them apply a prior-commitment rule — an expense incurred before filing is excluded. That verification is what opens step 5 of the audit.
See every programFrequently asked questions
An AI audit is a diagnostic that establishes what a company can realistically automate, predict or classify with artificial intelligence, based on the systems, data and processes it already has. It has five parts: mapping the technology environment, identifying the bottlenecks, qualifying the points where AI can fit, a sequenced action plan, and a funding plan. It differs from a technology demonstration in that it starts from the constraints of the company rather than the capabilities of a tool.
With an inventory of the data you already hold, not with the choice of a technology. Most AI projects that fail in a company fail because the necessary data was not being collected, was not kept long enough, or was scattered across systems that do not communicate — a problem no model corrects. The right first question is therefore twofold: which problem costs the most today, and do we have the history to attack it?
No, but you need to know what you have. An AI audit takes data in the state it is in — spreadsheets and manual entry included — and determines which of it is usable as is, which requires preparation work, and which does not yet exist and will have to be collected before a use case becomes possible. That data preparation work is itself an eligible expense under several funding programs.
Yes, in several cases. Fees for a diagnostic carried out by an independent third party are an eligible expense under sub-stream 1C of Investissement Québec’s ESSOR program, under PROMPT’s technology requirements document, and under the support vouchers offered by some cities including Québec City. Two conditions come up everywhere: the engagement must go to a third party — your own team’s hours are not eligible, even valued at cost — and the application must be filed before the expense is incurred.
Yes, for most of the programs listed. The prior-commitment rule excludes any expense incurred or invoiced before the filing date, and it is the leading cause of refusal on ESSOR sub-stream 1C. In practice the application is built on the service offer: an engagement already under way is no longer fundable, and there is no retroactive fix.
An integrator is paid by the solution it installs: its diagnosis leads naturally to its catalogue. Hubvention sells no software and installs nothing, so the plan that comes out of the audit can be executed by your own teams or by the integrator of your choice. Added to that is the part few firms cover: preparing the public funding applications attached to the project, from filing through to payment.