Why I Left Corporate Real Estate to Bet Everything on AI

The morning after the Intrum acquisition closed, I sat in my office in Germany staring at a 47-page integration playbook. Page 12 outlined the new reporting requirements. Daily activity logs. Weekly pipeline reviews. Monthly performance dashboards with drill-down capability to individual call records.

I counted the required fields per broker per day. Thirty-eight. Thirty-eight fields that had to be filled in before the system would let them move to the next screen.

I’d just sold a company I’d built over nearly 15 years. I should have felt relief. Instead, I felt nauseous. Because I recognized what was coming — I’d built a lighter version of it myself. And I knew what it would do to the people I’d spent years recruiting and training.

What corporate ownership does to a brokerage

I won’t pretend the acquisition was a mistake. It wasn’t. Selling Assetgate to Intrum was a legitimate business decision. They’re a serious company. The deal was fair. But the moment a brokerage becomes a line item inside a financial services corporation, the priorities invert.

A standalone brokerage exists to close deals. A corporate-owned brokerage exists to produce reports that prove it’s closing deals. The distinction sounds small. It’s everything.

Within six months of the acquisition, my top broker — 18 deals a year, built relationships that lasted decades — dropped to 11. Not because he forgot how to sell. Because his Mondays and Tuesdays were now consumed by system entries and status meetings. Three days a week to sell. Five days a week to document.

Within a year, three of my best people had left. Not for competitors. Not for more money. They left the industry entirely. One became a consultant. One opened a restaurant. One told me, standing in the parking lot on his last day, “Jorg, I became a broker because I love talking to people. I haven’t talked to a client in three weeks.”

That parking lot conversation changed my life. Not immediately — it took a few more months of watching the machine grind before I could articulate why. But the seed was planted right there, between two cars, on a Thursday afternoon.

The moment I decided to leave

It wasn’t dramatic. There was no board meeting blow-up, no public resignation. It was a Tuesday evening in my home office, reviewing the quarterly numbers.

Revenue was down 22% from the year before the acquisition. Not because the market was down — the market was fine. Because the brokers were slower. The pipeline was thinner. The energy was different. People were going through the motions, checking the boxes, filing the reports, and somehow doing less actual work than ever.

I opened a spreadsheet I’d been maintaining privately — tracking how my brokers spent their time before and after the integration. Before: roughly 40% of their week on client-facing activities. After: 23%. The other 77% had been swallowed by corporate reporting infrastructure.

I stared at those numbers and thought: I built something good. I hired good people. I created an environment where closers could close. And now I’m watching a reporting system eat it alive.

The next morning, I started planning my exit.

Why AI and not retirement

I could have walked away. The acquisition gave me options. I could have bought a boat, moved somewhere warm, and spent my days not thinking about CRMs and reporting requirements. People told me I should. “You’ve earned it,” they said. “Take a break.”

But here’s the thing about watching a system destroy good people for three decades — you can’t unknow it. Every broker I talked to, in every market, described the same frustration. Germany, Spain, Portugal, Latin America — different languages, different regulations, same complaint: “I don’t have time to do my actual job.”

And then AI happened. Not the buzzword version — not “AI-powered” stickers on the same old dashboards. Real AI. Models that could read context, understand intent, maintain conversations, anticipate needs. Models that could do the preparatory work my brokers were drowning in.

I saw the possibility immediately. Not because I’m a technologist — I’m not. I’m a real estate guy. But I spent 15 years watching talented people buried under admin, and suddenly a technology existed that could handle that admin without requiring anyone to type into a single field.

That’s not a nice feature. That’s a fundamental rewrite of how brokerages work.

So instead of a boat, I bet my exit money on building it.

Ten investors said no

I pitched Leon & Vera to ten PropTech investors. I explained the thesis: brokers are drowning in admin, AI can do the preparation work, the broker’s only job should be talking to clients and closing deals. Build for the closer, not the supervisor.

Every single one said no.

Their questions told me everything. “What analytics does the manager get?” “How do you track agent performance?” “What’s the dashboard like?” “How does the franchise owner monitor their team?”

Not one — not a single one — asked: “How much time does the broker save?” or “How many more deals can a broker close?”

They were thinking like the managers who buy software. Not like the brokers who suffer through it. They wanted the same surveillance tools wearing a new costume. AI-powered tracking instead of manual tracking. Smarter dashboards instead of dumb ones. Better visibility for supervisors.

I wasn’t pitching better visibility. I was pitching the elimination of the entire visibility tax. Stop tracking broker activity. Start enabling broker outcomes.

All 10 said no. I think that tells you everything about how the industry thinks about brokers.

Betting your own money changes how you build

When you raise venture capital, you answer to investors. Their metrics, their timelines, their assumptions about what real estate software should look like. Those assumptions are why PropTech keeps building the wrong thing.

When you bet your own money, you answer to one question: does this actually help the broker?

I don’t have a dashboard roadmap. I don’t have a feature matrix designed to impress franchise owners in demo rooms. I have one test: would the broker I lost in that parking lot — the one who hadn’t talked to a client in three weeks — would this product give him his job back?

That’s a different filter than “will this product sell to managers.” It produces fundamentally different software. Software that prepares instead of tracks. That delivers instead of requires. That works while the broker sleeps so they can hit the ground running at 8 AM instead of spending the first two hours feeding a machine.

The difference between investor money and your own money is simple: investor money optimizes for the buyer of the software. Your own money optimizes for the user.

What I’d tell the version of me sitting in that Stockholm demo room

If I could go back to 2019, watching that CRM demo, knowing what I know now, I’d stand up and ask one thing:

“Show me the screen where the broker doesn’t have to be here.”

Because that’s the whole point. The best tool for a broker is the one they barely touch. The one that finds the next lead, prepares the conversation, delivers the context — all before the broker opens their eyes in the morning.

Not a smarter dashboard. Not a better reporting interface. Not “AI-powered” anything bolted onto the same architecture. An entirely different relationship between a broker and their tools. One where the tool does the work and the broker does the closing.

I left corporate real estate because I watched reporting infrastructure crush the best closers I’d ever hired. I bet everything on AI because for the first time in 30 years, the technology exists to give brokers back the 72% of their time that surveillance software stole.

Ten investors thought I was crazy. They might be right about the risk. But they’re wrong about the brokers. Brokers don’t want better dashboards. They want their time back. They want to do the job they were hired for.

I couldn’t fix that from inside a corporation. So I left. And I bet everything I had on the belief that the next era of real estate technology starts with a simple question: what would you build if the broker’s time was the only thing that mattered?

I’m still answering that question every day. But at least now I’m asking the right one.