About Working Fluency AI

Technology is only useful when you can work with it.

Working Fluency AI is my practice, based in Portland, Oregon.

Matthew Westerbeck leaning in beside a bootcamp attendee as she works on her laptop

Matthew Westerbeck, founder

I come to this through the work.

Before Working Fluency AI, I spent ten years in nonprofits and federal contracting, leading programs, developing teams, managing contracts and grant compliance, and building data systems.

I served as a director of a refugee resettlement program in Oregon, and then spent nearly three years as a program manager with a federal contractor. There, I ran a six-month federal humanitarian response program serving Afghan refugees, where I was responsible for its team of more than 400 people, its $30 million operating budget, and more than $12 million in emergency direct assistance distributed in four months. I’ve worked inside the constraints that come with a caseload, a tight timeline, and responsibilities that don’t pause while you learn new software.

That experience shapes what I want AI to make possible. When the administrative work takes less time, there’s more room to listen carefully, build trust, and give a difficult decision the attention it deserves, and those human parts of the work are why efficiency matters to me.

My goal is twofold: help you achieve real productivity and efficiency gains now, and develop the skills to keep improving how you work.

Claude Certified Associate - Foundations badgeClaude Certified Associate - FoundationsIssued by Anthropic. Verify on Credly

The name

What working fluency means.

Working fluency means putting AI to work, judging the result, and adapting when the work or the tool changes. My work delivers exactly that: tools set up on your real work, and the judgment to run them well after I leave.

Fluency grows through use. As you gain confidence, you can experiment, spot bottlenecks, and tackle problems you hadn’t thought to bring to AI before.

Practice

Learn on real work.

Your files, your tools, and a task worth doing, so learning and useful progress happen together.

Judgment

Keep the review yours.

Give the tools context and standards. Check the output, trace its sources, and decide what is ready to use.

Independence

Build on what you learn.

Leave able to run the workflow, improve it, and recognize the next opportunity to make your work easier.

Matthew Westerbeck and Shemia Fagan co-teaching the TLU Claude Bootcamp

The teaching partnership

Teaching lawyers from working experience.

I teach workshops for small businesses and nonprofits and work with clients one-on-one on coaching and implementation. For lawyers and legal staff, Shemia Fagan and I teach the Claude Bootcamp, which firms can book directly with me; Trial Lawyers University hosted our 2026 classes.

Shemia is a trial lawyer and Managing Partner at HKM Employment Attorneys, where she runs the firm’s Boise, Oakland, Sacramento, San Francisco, and San Jose offices. She represents employees in cases involving discrimination, wrongful termination, harassment, retaliation, and wage theft, and she teaches from the systems she uses in her own practice.

A good first step

Start with a conversation.

Bring a question, a bottleneck, or an idea for doing things differently. We’ll work out where AI can help and what kind of support makes sense.

The first call is free, on Zoom, and needs no preparation. For on-site work, I travel anywhere in North America.