A few years ago I started taking on consulting work alongside my day job, first with paid sessions at IVADO’s MLOps training in Montreal, then more structured engagements from there. The common thread: helping teams do more with their data and ship “AI” that actually works in the real world.

Worth being upfront: this is a side activity. I have a 9-to-5 I enjoy and have no plans to leave. Maybe that can change down the road but for now, I keep the number of engagements small, which means each one gets proper attention. Availability is real, and I’ll always be honest about it from the start.

I keep things focused on four areas where I have real experience.


What I Can Help With

Web Scraping Systems

Systems

Designing and operating web scraping infrastructure that holds up in production. I've run these systems for years — as side projects and inside Ubisoft to support client work. I can help you think through architecture, robustness, long-term maintainability and legal limitation before the first headache hits.

Python Scrapy Playwright AWS

AI Operationalization

MLOps

Getting ML from a notebook into production is where most projects either succeed or quietly die. With nearly 10 years in this space, including 5 years building an in-house ML platform from scratch. I've seen what works and what wastes months. Whether you're shipping your first pipeline or scaling an existing one, I can help.

MLflow Databricks Feature Stores Model Monitoring

Data Investigation

Analysis

Got a messy dataset and no idea where to start? I spent years at EDF digging into everything from customer call transcripts to power plant turbine data. I'll surface the signal, prototype a proof of concept from an idea, and give you a clear picture of what's possible.

Python Exploratory Analysis POC Development NLP

Talks & Events

Speaking

Sharing knowledge is its own craft and a different exercise from building pipelines. I've spoken at events like IVADO's MLOps training in Montreal, pydata/mlops events, and I've been on both sides of the table: as a speaker and as part of the organizing team of pydata/mlops community MTL communities. If you need a technical speaker or help curating a lineup, let's talk.

MLOps Data Science Event Organization

How It Works

Getting started is simple by design. No lengthy proposals, no unnecessary overhead — just a clear three-step process from first contact to focused work.

01

Discovery Call

A structured 45-minute conversation to map out the full picture of your project — covering the 5 Ws: what you need, why it matters, who's involved, when you need it, and where the blockers are. The goal is mutual clarity on scope and fit. No commitment on either side.

02

Scope & Proposal

If the call confirms a clear fit, I'll follow up with a short written proposal: what we'll work on, the approach, an estimated number of hours, and the rate. Nothing bloated — just enough to align on expectations before any work starts.

03

Hourly Engagement

Once the proposal is agreed, we get to work. The engagement is hourly, scoped to your needs, and tracked transparently. No retainer, no lock-in — just focused work with a clear end state in mind.


Sounds like a fit?

Reach out on LinkedIn (with a nice note if we are not connected yet) and let's figure out what we can do together.

Get in Touch on LinkedIn