GoodMind AI helps Quebec and Canadian small and medium businesses integrate artificial intelligence where it solves real problems — so your teams can focus on their core mission. And when AI is not the right answer, we tell you.
Why GoodMind AI
"I founded GoodMind AI on a simple conviction: your AI projects deserve the same degree of rigour as the systems built for medical research and clinical trials."
Before GoodMind AI, I spent fifteen years building software in contexts where precision truly matters: medical imaging, surgical navigation, MRI-compatible robotics, deep learning and clinical trials.
At Harvard Medical School, I learned that a good system isn't the one that impresses in a demo: it's the one that stays understandable, traceable and robust when it meets real-world cases. My PhD in deep learning then taught me to treat models as imperfect statistical tools — useful when properly evaluated, dangerous when they replace judgment.
With GoodMind AI, I bring that discipline to Quebec and Canadian SMBs: listen before prescribing, say no when AI is not the right answer, and build systems simple enough to be used for a long time. I'm convinced that's the condition for staying globally competitive without losing sight of your mission.
Research engineer: medical imaging, surgical navigation, robotic integration and open-source software.
Deep learning for brain image analysis, efficient algorithms for massive datasets, scientific publications.
Research scientist: state-of-the-art AI systems for large-scale brain image processing in clinical trials for neurodegenerative diseases.
Founder: turning the recurring problems of SMBs into useful, maintainable systems.
01
Systems are designed to be tested, explained, documented and maintained.
02
AI is proposed only when it brings more value than simple automation.
03
Data, access, costs and Law 25 requirements are addressed from the diagnostic onward.
04
You get readable deliverables: recommendations, risks, costs, limits and next steps.
Real examples
Every project starts from a daily pain point — not from a technology. Here is what that looks like in four sectors.
01
"We spend hours digging through old files to find the right clause."
Your files answer back. An assistant retrieves precedents from past files, drafts standard clauses and prepares deed drafts — citing its sources.
The notary focuses on analysis, not paperwork.
02
"Incoming requests pile up and we can't call everyone back."
A front desk that never sleeps. A dedicated assistant triages requests, proposes first appointments, sends reminders and collects intake forms — in full compliance with Law 25.
The team gets time back for patients, not logistics.
03
"Reconciling client transactions by hand is slow and error-prone."
Your data falls into place. A system gathers and structures transactions, detects anomalies and prepares draft GST/QST reports ready for review.
The accountant reviews and advises instead of re-keying data.
04
"Preparing a bid from a call for tenders takes us days."
A first bid in minutes. AI prepares a first draft from the call for tenders, drawing on your past quotes, supplier prices and RBQ standards.
You start from a solid draft and keep full control over the final price.
Our approach
No big contract to sign blind: every phase produces deliverables that belong to you, and you decide what comes next.
Understand the problem, the available data, the feasibility and the potential value. Together we define the metrics that will decide whether the project is a success — before writing a single line of code.
Deliverable: a readable diagnostic report — recommendations, risks, estimated costs, limits and next steps.
A prototype on a limited scope, tested with your real data. Your team tries it, gives feedback, and you decide with full knowledge whether to move forward.
Deliverable: a working prototype and an honest evaluation of its performance.
Production-grade development and deployment in your environment, with optional training for your employees to ensure lasting adoption.
Deliverable: a deployed, documented system — and a team that knows how to use it.
Maintenance, optimization, cost tracking and new features, at the pace of your needs.
Deliverable: a system that keeps performing, with costs under control.
At the end of each phase, you are free to stop, continue with us, or take your deliverables to another partner. Everything we produce belongs to you: code, documentation, results.
Data sovereignty
Protecting personal information is not an afterthought. Every project starts with a simple question: where does your data go, and who can access it?
Blog
Canada is putting half a billion dollars on the table so that SMBs stop watching AI from the sidelines. Here's what LIFT offers, who qualifies, and how to make the most of it.
Read articleThe most profitable question to ask before any AI project fits in one sentence: what problem are we trying to solve?
Read articleHanding data to an AI tool is, first and foremost, handing over data. Law 25 doesn't forbid it — it requires you to know what you're doing.
Read articleTwo minutes to assess your project's readiness — or one email to talk it through directly. Either way, you leave with an honest answer.