Pat Veilleux

AI for people
who do things.

I don't teach theory. I teach what I've built, tested, and used — on real work, for real clients — then stripped down so you can actually use it.

Based in Costa Rica
Background Lawyer · Operator · Builder

I build the thing first.
Then I teach it.

I spent years as senior counsel at Shopify, then built a practice advising companies across nuclear energy, healthcare AI, and fintech. Along the way I became obsessed with a single question: what happens when you apply AI to the most judgment-intensive work there is? If it works for legal — it works for anything.

The Judgment File.

The concept at the centre of everything I teach. Simple in principle, powerful in practice.

Your AI outputs are only as good as the
context you give it.

Most people use AI like a search engine — they type a question and hope for a good answer. The result is generic output that doesn't know who they are, how they think, or what they're actually trying to accomplish.

The Judgment File changes that. It's a persistent context document — built once, refined over time — that tells your AI tool everything it needs to produce output that sounds like you, thinks like you, and serves your actual situation. When your AI has your judgment baked in, it stops being a tool you prompt and starts being a system you trust.

Step one
Capture

Document who you are, how you think, what you're working on, and what good looks like in your world.

Step two
Distill

Strip it to the essential signals — the context that actually changes AI output, not everything you could say.

Step three
Deliver

Feed it systematically into your AI workflows. Watch the quality of your outputs change immediately.

Two levels.
Both matter.

Most people try to run before they can walk with AI. Here's how I actually think about where you should start — and where you should go.

01
Level one

Do your own work faster

Before you think about your team, you need to feel it yourself. This level is about personal productivity — using AI to compress the time between a task landing and a high-quality output leaving your desk. Drafting, reviewing, researching, synthesizing. The goal is to make your own judgment faster to deploy, not to replace it.

02
Level two

Replicate your judgment across your team

Once you've internalized what good AI-augmented work looks like, you can systematize it. Workflows, shared context, repeatable prompts, team standards. This is where AI stops being a personal productivity tool and starts becoming an operational multiplier — elevating everyone's output, not just yours.

Pat Veilleux

Lawyer by training.
Operator in practice.

I'm a Canadian lawyer with almost 20 years advising companies on their hardest decisions — from Shopify to nuclear energy startups to healthcare AI platforms. My bias is toward action. When I want to understand something, I build with it. Every workflow I teach, I've run live. Every framework I share, I've stress-tested on real work.

GC-level work across nuclear, healthcare AI, and fintech
Former Senior Counsel, Shopify — and Maven Clinic, Oklo
Almost 20 years advising companies on their hardest decisions
Work with Pat
If you do judgment-intensive work,
this is for you.

Workshops for lawyers, legal teams, and the professionals who sit alongside them. Practical, built from real work, no fluff.

Follow on LinkedIn