Rocketable Is Buying Software Companies and Firing Everyone Except the AI

The Macro: The Roll-Up Model Gets an AI Upgrade

Software roll-ups aren’t new. Constellation Software has been buying vertical SaaS companies since 2006 and now has a market cap north of $70 billion. Andrew Wilkinson built Tiny Capital on a similar premise: buy boring, profitable internet businesses and hold them forever. Private equity firms have been rolling up SaaS for years, cutting costs and optimizing margins through the standard playbook of consolidation and layoffs.

What’s different now is that AI can plausibly replace entire job functions that roll-up operators previously had to staff. Customer support. Basic development and bug fixes. Content updates. Account management. Marketing. If you’re buying a $2 million ARR SaaS product and the cost structure is $1.5 million in human labor, the math changes dramatically when AI agents can handle 80% of that work. Your margins go from 25% to something approaching 80%.

The question is whether this actually works in practice. There’s a massive gap between “AI can draft a customer support email” and “AI can run customer success for a B2B SaaS product without anyone churning.” Anyone who’s deployed AI in production knows the difference. But the economic incentive to try is enormous, and someone was going to build a company around this thesis. Several someones, actually. The race is on.

The Micro: A Self-Driving Car Guy Who Wants to Drive Companies

Alan Wells is running Rocketable as a solo founder, which is fitting for a company whose entire thesis is that you don’t need large teams. His background is exactly what you’d want for this kind of bet: he worked on robotaxis at Cruise, self-driving trucks at Uber, and co-founded Tule (YC S14), an AI-powered irrigation company. He holds five patents in AI automation. The man has spent a decade building systems that replace human decision-making with machine decision-making. Now he’s applying that to entire companies.

The model is straightforward. Acquire existing software products with real revenue and real customers. Then systematically replace the human org chart with AI agents. Rocketable is building internal tooling to make this repeatable across acquisitions. They’re a two-person team working remotely, part of YC’s Winter 2025 batch.

I find the framing honest in a way that most AI companies aren’t. They’re not calling it “augmentation” or “empowerment.” They’re calling it replacement. The AI Maximalist Software Holding Company. That’s either refreshingly direct or a future PR problem, depending on how the next few years play out.

The Verdict

This is one of those ideas that sounds crazy until you run the numbers. A portfolio of ten small SaaS products doing $1-3 million each, operating at 80%+ margins because AI handles most of the work? That’s a $10-30 million revenue business with maybe $20-24 million in profit. With two humans. The economics are obscene if it works.

The risk is quality degradation. SaaS customers pay for reliability and responsiveness. If an AI agent mishandles an enterprise customer’s support ticket, or ships a bug that breaks a workflow, or sends a tone-deaf marketing email, the churn compounds fast. Small SaaS products often survive on the strength of personal relationships between the founder and key accounts. Replacing that with AI is theoretically possible but practically unproven at scale.

There’s also a competitive concern. Tiny Capital, Constellation Software, and dozens of PE firms are all looking at the same acquisition targets. If Rocketable’s edge is AI operations, that edge only lasts until everyone else adopts similar tools. And they will.

In 30 days, I’d want to know how many acquisitions they’ve closed and what the handover from human team to AI agents actually looks like. At 60 days, customer retention numbers will tell the real story. By 90 days, if they can show a portfolio company running profitably with near-zero human involvement and no spike in churn, that’s proof of concept that will attract serious capital. I’m skeptical but intrigued. The thesis is sound. The execution is the whole game.