AI-generated
Adrian Föhl
AI Strategy & Leadership · DüsseldorfAI Strategy & Leadership · Currently building the AI function for 1,300 people in healthcare · Based in Düsseldorf, Germany
I know the technology. But the hard part is never the technology. It’s getting an organization to actually change.
Half Filipino, raised in Germany. I grew up switching between languages, cultures, and ways of seeing the world. That taught me more about building for people than any CS degree.
I left Germany at 19 to backpack Australia, and kept going from there. An internship in Singapore, a semester in Sydney, always drawn to whatever felt slightly too early. I’ve been in tech since 2011, working through every hype cycle from mobile apps to AR to AI. The thread that connects all of it: getting organizations to adopt new technology before they feel ready.
Click to explore each chapter
Everyone wanted mobile apps. Nobody knew why. At T-Systems, I learned the hardest part: getting teams to agree on the problem before touching the code.
Things I’ve shipped.
What actually happened.
AI for a 1,300-person corporate group
Not a lighthouse demo. Adoption across the whole group.
Multiple subsidiaries, each with its own management, works council, and compliance requirements. A showcase project would have died in there. So: built a small AI team, developed a company-wide AI policy together, trained over 200 employees, from teams to managing directors. In early surveys, people report saving time in their daily work.
What I learned: The AI policy was written in weeks. Preparing the organization for it took months.
Pitching the AI pivot at a Fraunhofer spin-off
Didn’t pitch AI. Took work off their plate.
The company did smart city consulting, nobody was thinking about AI. Instead of selling tools, I automated the skeptical team’s most annoying tasks. Their own first time saved convinced them more than any slide. Plus exchange with over 20 municipalities, workshops, webinars. Started with a pitch to leadership; within a year, from zero to first delivered projects and a new business area.
What I learned: Seeing the wave early doesn’t matter if you can’t bring people with you.
A personal AI agent team I use daily
Stopped thinking in roles. Started thinking in handoffs.
I built a team of named agents: one researches, one pushes back, one plans. It worked, but the org-chart was just scaffolding. What lasts, and runs daily, is the layer underneath: orchestration, clean handoffs, research before conclusions, a memory in files that gets pulled in when it counts. It’s how this site, my research, and my writing get made.
What I learned: The org-chart was scaffolding to climb. Not the building.
What I believe.
What’s left when the hype fades.
Technology never fails because of the technology.
I've watched it happen three times. Mobile apps that died because the culture wasn't ready for transparency. AR that worked in the lab but not in the boardroom. AI that gets blocked by processes people have followed for twenty years. Every time, the real blocker was something non-technical: regulation, habit, fear. The teams that win figure out the human problem first.
Every assumption has an expiration date.
What was true about AI six months ago might already be wrong. Models get cheaper, capabilities shift, entire approaches become obsolete overnight. If you're not willing to constantly question what you know, you'll build on foundations that are already cracking.
AI should work autonomously. But never be a black box.
The real value of AI is when it handles work without you having to do every step yourself. But autonomous doesn't mean opaque. You should always be able to understand how a result was reached. Especially in regulated environments, trust comes from transparency, not from hiding the AI behind a button.
Building with AI
Honest write-ups of what I make with it, and what I get wrong along the way.
Let’s talk.
Anyone can pitch AI adoption. Making it actually work is the real job. I read every message.