How a 5,000-Person Multinational Trained Its Leaders to Build with Claude Code

What we can tell you is this: they operate in 27 countries, employ around 5,000 people, and work across some of the most data-intensive industries in the world. Environmental monitoring. Finance. Infrastructure. The kind of work where the gap between raw data and actionable insight has historically required significant vendor investment and long timelines.

They came to us because they had seen what Claude Code could do. They wanted to know if their people could learn to use it. Not their IT department. Not their engineers. Their leaders.

The training was led by Dr. William Hahn, an AI scientist and researcher with over a decade of experience as a university professor. Dr. Hahn earned his PhD in Machine Perception at Florida Atlantic University, where his doctoral research focused on complex systems and brain sciences. He has served as Co-Director of the Machine Perception and Cognitive Robotics Laboratory at FAU since 2011, one of the longest-running applied AI research programs in the region. He is not a consultant who learned Claude Code last year. He is a scientist who has watched this technology develop from inside a university research environment and knows exactly what it can do in the hands of a business team.

Cohort One: Colorado

The first cohort ran in Colorado. Nine participants. Four days. Nine to five, with a break for lunch, in a professional conference room.

The participants were not developers. They were the kind of people who run departments, manage teams, and make decisions about where the company invests. The head of marketing. Regional directors. Senior managers. People who had never written a line of code and had no particular reason to believe they were about to start.

The first session was designed around a single goal: deliver a moment that breaks assumptions about what AI can and cannot do, within the first twenty minutes.

It worked.

Within the first half hour, participants had built working applications running on each other's phones. Not mockups. Not demos prepared in advance. Things they had built themselves, from scratch, that moment, that were live on the web and accessible by QR code. The reaction was consistent across all nine people: they had not believed this was possible without a development team, months of work, and a budget they would have had to justify to a committee.

By day three, participants were building tools relevant to their actual work. One participant used environmental monitoring data to build a live dashboard tracking conditions across a region. Another worked with financial datasets to build a reporting tool that had been on the company's internal wish list for over a year. The tools were rough in places. They were also real.

Day four was consolidation. Every participant left with a working project they had built themselves and a documented playbook for continuing to develop independently after the training ended.

The client liked what they saw. They booked a second cohort. In London.

Cohort Two: London

The second engagement ran in London. Fifteen participants. Same format: four days, nine to five, hotel conference room, leaders only.

The London cohort had one significant change from Colorado: a web-based training portal that eliminated the hardware and setup friction from the first engagement. In Colorado, the requirement had been a brand-new laptop, configured from scratch on day one. That process worked, but it consumed time and created friction that had nothing to do with Claude Code.

The portal changed that. Participants opened a browser, clicked, and had a full Claude Code environment running on whatever device they had brought. Windows laptops. Company machines. Any device. Zero install. Zero setup. When they hit enter on a project, it deployed live to the web and a QR code appeared. They scanned it. Their application was on their phone. The entire loop, from opening the browser to seeing a working application on a device, closed within twenty minutes on day one.

Within the first hour, participants were asking whether they could keep using the platform after the training ended. That question had not been on the agenda. It became the first indication that the portal itself had value beyond the training week.

The London cohort went deeper on data. Participants worked with real datasets relevant to their fields: environmental sensor data, groundwater monitoring, forest conditions, financial reporting. They built dashboards. They built APIs. One participant rebuilt a piece of internal software the company had been stuck with for years because the original vendor had gone out of business and the bugs were no longer fixable. She rebuilt it during the training week. It worked better than the original.

One session produced something that still stands as a clear demonstration of what this technology can do. A participant asked Claude Code to build a live sensor network using all fifteen phones in the room as microphones, running real-time audio classification across every device simultaneously, with results displayed on the main screen. One sentence of instruction. It ran on the first attempt. The room went quiet. Then everyone started talking at once.

That is the experience the training is designed to deliver: the moment when the gap between what people thought was possible and what is actually possible closes in real time.

What the Client Said After

The CEO visited the London cohort. He watched participants present what they had built during the week. After the session, he told us something that has shaped how we talk about this training.

The fifteen people in that room, he said, collectively influence approximately one thousand employees across the organization.

He was not talking about a technology rollout. He was talking about culture. The people who go through this training do not come back to their teams with a new software tool. They come back with a different understanding of what is possible. They become the people in the room who say "we can build that" instead of "we would need to hire someone for that" or "we would need to budget for a vendor." That shift, multiplied across the people those fifteen leaders influence, is what the CEO was describing.

One week. Fifteen people. One thousand downstream.

What This Means for Your Organization

This training is not right for every company. It is right for organizations that have real operational problems software could solve, but lack the development resources to address them at the pace the business requires. Organizations where leadership understands that AI fluency is becoming a competitive requirement, not a nice-to-have. Organizations willing to invest in building genuine internal capability rather than outsourcing every technical task indefinitely.

If those conditions are present, a four-day on-site intensive led by a working AI scientist can compress months of organic trial-and-error into a week of structured, expert-guided development.

The client we trained operates in 27 countries. They brought us back for a second cohort in a different country because the first one worked.

That is the only case study we need.