Today’s LLMs are exceptional at predicting the next word. Some are even excellent chess players. But chess is a closed system with fixed rules and a finite board. Predicting the ripple effects of a real business decision… how a pricing change plays out five moves later, after competitors react and customers adjust… is a different kind of problem entirely.

Skyfall AI, led by Sam Pasupalak, Kaheer Suleman, and Sumit Pasupalak, is on a mission to build the first Enterprise Super Intelligence. While most of the industry is racing to build the next virtual software engineer or sales agent, Skyfall is going after something harder: a virtual CEO. The neolab is building an Enterprise World Model designed to simulate the actual consequences of a business decision, not just describe it. Where an LLM generates text, Skyfall’s models are working towards forecasting the next state of an organization and the world it operates in. That’s a much harder problem, and a more useful one.

I’ve met enough founders to know the difference between someone chasing a trend and someone who’s been circling the same problem for a decade. Sam and Kaheer built Maluuba, one of the earliest deep learning labs, working alongside people like Yoshua Bengio and Richard Sutton years before the rest of the world caught up to reinforcement learning. Microsoft acquired Maluuba in 2017. Most founders would call that the finish line, but Sam is incapable of thinking small or sitting on the sidelines. He came back for another round, to build a lab to chase a bigger problem, with the conviction of someone who never stopped thinking about it.

Skyfall’s first product, Morpheus, is a continual reinforcement learning platform for AI researchers that lets agents learn inside persistent, non-stationary environments – starting with warehouse inbound and outbound management – without relying on explicit task labels or scripted curricula. It’s also the system Skyfall’s own research team uses to train their models, and the foundation for an agentic suite they plan to release in the coming months to automate dozens of enterprise tasks for CEOs and entrepreneurs.

They’re also about to test the thesis in the most literal way possible: acquiring a small SaaS company and running it with virtually no human hands on the wheel. Humans sign the legal paperwork, set up the initial systems, file the taxes. Everything else – pricing, growth, operations – will be on the AI to execute. The goal is to double revenue in six months. Most AI companies prove themselves with a benchmark or a demo. Skyfall is proposing to prove itself by actually running a business and living with the consequences if it doesn’t work.

Authentic. Relentless. Inevitable. Those are the three words I keep coming back to when I think about why we backed Sam and his cofounders at inception. Back then, I saw in Sam a founder with the outsized ambition, drive, and expertise necessary to build something of consequence in this moment. His obsessive, grounded conviction is exactly what it takes right now. It’s been a privilege to watch Sam’s big vision and maniacal focus lead to Skyfall.

Congratulations to Sam, Kaheer, Sumit, and the entire Skyfall team on coming out of stealth. You can learn more in Forbes’s deep dive. We’re excited to be along for it.