What You’ll Learn

James went to a startup pitchfest in Ottawa where founder after founder talked about launching in months — minimum viable products, six-month roadmaps, the works. In this live session he does the opposite: he builds a real, sellable digital product in about an hour, start to finish, on camera. This walkthrough breaks down how he did it and, more importantly, the repeatable method underneath it.

By the end you’ll understand what counts as a “digital product” (it’s not another PDF), how to spot one hiding in your own routine, and how to build it — plus a reusable playbook — so the next one takes a single conversation instead of a week.

The Big Lesson: Capture What You Already Do

Here’s the shift James is pointing at. With today’s powerful AI, the scarce resource isn’t the technology — it’s good ideas and context for the models to work on. Most people hear “digital product” and immediately think ebook or PDF. That’s not it.

“The big lesson is to capture what you’re already doing on a regular basis that you know provides value. Get something up and running you can use, then start testing it as something you might monetize.”

💡 In Plain English: the product you should build is the boring thing you do every week without thinking. If it saves you time, it’ll save your audience time — and they’ll pay for that.

His example came straight from his customers: coaches told him they spend 45 minutes to an hour prepping for each one-on-one call. Erase most of that time, and would a coach pay for it? He knew plenty who would — so that became the product.

Step 1: Point Your Orchestrator at a Clear Idea

James opens Claude with the campus operating system in the working folder and talks to Dean, his chief orchestration officer. Because Dean already holds his foundational memory — what the business is, what it sells, its tools, the last several days of work, and current goals — James doesn’t have to re-explain anything. He just says: let’s build a new agent team that gets me prepped for any interaction with real people.

Step 2: Use the Right Model for the Job

For this build he deliberately switches to Fable 5 (on high effort) rather than Sonnet, because it’s stronger at work that needs more thinking, nuance, and intent. This is the model-selection discipline in action — match the model to the difficulty of the task.

Step 3: Build the Product AND the Playbook at the Same Time

This is the part that compounds. While Dean builds the product, James has him capture a reusable playbook for doing it again. The goal: next time he just says “I have an idea for a product,” and the system writes the plan (a PRD), checks it with him, then assigns the pieces to the right employees and builds it.

The product itself is the Prep Agents Team — a plugin that adds to his campus operating system. It’s persona-adaptive session prep for anyone who presents, teaches, coaches, or sells: onboarding identifies who you are, then every run adapts across two pipelines (group presentations and one-to-one calls) and two passes (before and after a session). The outputs include agendas, decks, handouts, cue cards, call briefs, recaps, and follow-ups — all in your voice and brand.

Step 4: Install and Warm-Test It

About an hour and 25 minutes in, the plugin is packaged into zip files with a test script and conformance checks. James installs it into his campus (a “warm” install, since the operating system is already in the folder) and runs the onboarding — which recognizes everything it already knows about him and his business, so setup is nearly instant.

Step 5: Let the System Handle the Go-to-Market

Building the product is only half of it. The same operating system then produces the launch: email marketing, social posts, community announcements, product descriptions, payment setup, and the sales page. James points to the product he shipped the day before — Scout, a 10-agent team — where a follow-on playbook wrote the entire sales page and created the product in his shopping cart with pricing already set.

The Takeaway

You don’t need six months and a minimum viable product. If you have a task you already repeat that clearly delivers value, you can capture it as a digital product, build it and its playbook in an afternoon, and let your operating system handle the launch. Start one step earlier than James did here and it’s still a one-day job, not a one-quarter project. If you want the system he’s using, the campus operating system is at trainingsites.io/os.

Teach more, and let the agents do the rest.

This tutorial is a recap of a live build session.