Start with the job
Understand the person, the decision, and the failure cost before choosing a model or framework.
About / Montréal
I learned the problem before I learned the stack.
A
From the front line to the product layer.
My route into engineering started with people: customers, teams, campaigns, operational bottlenecks, and the gap between what software promised and what the work actually required.
Marketing taught me to make an outcome legible. Operations taught me that details, handoffs, and failure modes matter. Client work taught me to listen before prescribing. Software and applied AI gave me a way to turn those lessons into systems.
Today I work at the intersection of product engineering and AI: native local-first applications, agent workflows with real boundaries, evaluation harnesses, and automations that leave evidence behind.
Working philosophy / 01–04
Bound the system
Declare permissions, artifacts, review points, and what “done” means.
Measure honestly
Keep fixtures, definitions, hardware, sample sizes, and limitations beside the result.
Ship the whole path
Architecture matters, but so do packaging, permissions, recovery, documentation, and the last interaction.
Focus
Local AI · agent systems · evaluation · product automation
Languages
English, French, and Spanish
English is the primary technical language. Professional collaboration is also available in French and Spanish.