Fort Lauderdale
Florida, United States
The US office, and the one most engagements run out of.
Open in Maps ↗Montreal
An AI consultant
in Montreal.
An office in the city and twenty-five years of building software organizations behind it. I take the Chief AI Officer seat, write the plan, and stay to build it. Montreal is a city with more AI research per square kilometre than almost anywhere, and a lot of companies here still have nobody accountable for using any of it.
Montreal officeQuebec and across CanadaPublished pricingMid-market, typically $10M–$50M
Oshri CohenAI Consultant & Fractional CAIO
The short answer
Research nearby.
Adoption is another matter.
Montreal has an unusual amount of AI talent for a city its size, and almost none of that concentration reaches the mid-market company running a distribution business in Ville Saint-Laurent or a services group on the South Shore. Those companies do not need a research lab. They need somebody senior to decide which two workflows are worth automating, what it costs to run, and who owns it afterwards.
That is the work. I take the Chief AI Officer seat, fractionally or on an interim basis, and own the outcome from the plan through to production. The engagement is described in full on the AI consulting page, and the rates are published rather than discovered on the third call.
I keep offices in Montreal and in Fort Lauderdale, both on Eastern Time. Most of my work is with companies in the United States and Canada, with the US as the larger market, and the Montreal office is how the Canadian side gets served properly rather than as an afterthought.
Also searched as: AI consultant near me Montreal, AI consulting firms Montreal, artificial intelligence consultant Quebec, consultant en intelligence artificielle Montréal, fractional AI leadership Canada.
Who calls
The Quebec companies
that reach out.
Mostly established businesses with real operations, rather than the startup layer the city is better known for.
✕
Manufacturers and distributors
Quoting that takes days, customer orders retyped into the ERP, and a catalogue that three systems describe differently.
✕
Transport and logistics
Carriers, brokers and 3PLs where the coordinator count rises in lockstep with volume because email is the integration layer.
✕
Companies with a US customer base
Serving American clients means answering American security questionnaires, and those now have AI sections nobody prepared for.
✕
Businesses with a privacy exposure
Quebec's privacy rules bite harder than most of North America, and automated decision-making is specifically in scope.
How I work here
Six ways in.
Nearly everyone starts with the diagnostic. Fixed fee, yours to keep, and it will tell you honestly whether you need me.
✳
AI Consulting
The full engagement: where AI pays, what to build, and the leadership to get it into production.
◆
Fractional Chief AI Officer
The seat, two days a week, with decision rights attached. Published pricing.
▲
AI Governance & Risk
Policy, data boundaries and review gates. The first thing to fix if your exposure is the urgent part.
●
AI Automation Consulting
Back-office and operations work: document intake, exception routing, systems that never connected.
■
AI Readiness Assessment
Whether your data, systems and team can support the ambition, scored before the budget is committed.
✦
Fractional & Interim CTO
When the problem is technology leadership broadly rather than AI specifically.
Where I work from
Montreal,
and Fort Lauderdale.
Two staffed offices on one clock, which is why coverage across both countries works without a handoff.
Montreal
Quebec, Canada
The Canadian office, and where the practice started in 2018.
Open in Maps ↗No street address is published yet, deliberately. When there is one it will appear here and on every other surface at once, because the offices are defined in one place and read everywhere.
This city exports AI researchers and imports AI slide decks. The gap in between is an operator problem.
Oshri Cohen
Common questions
From Quebec buyers.
Do you work with companies across Quebec and the rest of Canada?
Yes. The office is in Montreal and I work with companies across Quebec, Ontario and the rest of Canada, as well as throughout the United States. Engagements run remotely with travel for kickoffs and executive sessions. Everything runs on Eastern Time, which covers both countries inside a single working day.
How does Quebec privacy law affect an AI project?
It affects it early, which is the opposite of how most companies sequence it. Quebec's privacy regime places real obligations around personal information and gives individuals rights in relation to decisions made by automated means, so where data may go and what a system is allowed to decide alone are questions to settle before you build rather than during the security review. I build that into the governance work. For a binding read on your specific obligations, get counsel who practises in Quebec; I am not one.
My customers are American. Does that change anything?
It changes what you will be asked to prove. US enterprise buyers increasingly include AI questions in their security questionnaires, covering where data is processed, whether vendors may train on it, and how automated decisions are logged. Being able to answer those cleanly is a sales asset. Failing them quietly costs deals you never learn you lost, which is why governance work often pays for itself commercially rather than defensively.
Are your rates in Canadian or US dollars?
All published pricing is in US dollars: a fixed-fee diagnostic from $20,000, advisory at $10,000 a month, the fractional Chief AI Officer seat at $18,000 a month, and interim coverage at $30,000. For Canadian engagements the invoicing arrangement is straightforward to sort out, and I would rather quote one currency publicly than maintain two price lists that drift apart.
Is there an AI consultant near me in Montreal?
Several, and distance is the weakest filter you could apply. Ask what they would need to see in your data before committing to an outcome, what the system costs to run each month at your volume, and who owns it after they leave. Those answers tell you far more than a postal code. My pricing is published so you can decide whether a first conversation is worth anyone's time.
Montreal company,
AI on the agenda?
Tell me what is stuck: a pilot that will not ship, a security questionnaire you cannot answer, or a board asking what the plan is.