L">
👨‍💼 The Instructor

I'm Tyler Folkman — and I Build AI for a Living

I spent the past decade designing and deploying machine learning systems for Fortune 500 companies, healthcare organizations, and high-growth startups. I co-founded Zeff, an enterprise AI platform that was acquired by DataRobot.

Along the way, I realized something: most data science education focuses on theory while skipping the hard parts — how do you actually debug a failing model? How do you explain your results to a non-technical executive? How do you decide which algorithm to use when five might work?

Learning With Data is my answer to that gap. Everything I teach here is grounded in what I've learned building real systems — with real stakes.

MSc Computer Science PyTorch Certified 10+ Years in AI/ML Startup Founder (Acquired) Substack Top Writer
Tyler Folkman, Senior Data Scientist and Founder of Learning With Data, at his workspace
🎯 Our Mission

Democratizing Expert-Level Data Science Education

The gap between academia and industry is widening. Too many people learn algorithms from textbooks but can't navigate a real dataset. We exist to close that gap.

The Learning With Data workspace — a modern high-tech office where courses are built and tested

Practitioner-First

Every concept is taught by someone who's implemented it in a production environment — not just read about it.

Rigorously Honest

We show you when algorithms fail, when they're overkill, and what the industry actually uses day-to-day.

Community-Driven

Our newsletter and comment sections are places for genuine discussion — questions get real answers from Tyler himself.

📋 Editorial Policy

How We Create Content

Every article and course module on this platform is written, tested, and verified by Tyler Folkman personally. Code examples are executed in a live environment before publication.

We follow a three-stage process: (1) Draft based on direct professional experience, (2) Technical review where all code is run against current library versions, and (3) Clarity edit to ensure concepts are accessible to the stated audience level.

External sources — research papers, benchmark datasets, official documentation — are always cited directly. We do not publish content that we haven't personally validated.

If you find an error or an outdated code snippet, please contact us — we update content within 48 hours of a valid correction report.