How Non-Tech Founders Are Taking Over Enterprise AI Without a Line of Code

In a startup ecosystem obsessed with ‘builder’ narratives, Craft’s latest $15 million Series A is a quiet rebellion. The company, founded by Ilya Levtov—a Juilliard-trained cellist with zero coding experience—has built a supply chain intelligence platform that’s now in use at Fortune 500 manufacturers. This isn’t a fluke. It’s a signal.

Levtov’s journey underscores a tectonic shift in enterprise AI: the rise of ‘non-technical’ founders leveraging AI as a force multiplier for domain-specific problems. Craft’s platform, which ingests siloed supply chain data to predict disruptions, was built using no-code tools and off-the-shelf AI APIs. The result? A product that solves real pain points for procurement teams—without requiring a single engineer to write a line of code.

This trend is accelerating. Investors are increasingly backing founders with deep industry expertise over ‘technical’ co-founders. Why? Because the bottleneck in enterprise AI isn’t the model—it’s the data. Companies like Craft, which understand the nuances of supply chains, manufacturing, or logistics, can design AI systems that actually move the needle. Technical debt is being replaced by domain fluency.

The implications are profound. For one, it democratizes AI entrepreneurship. A former supply chain manager, a healthcare administrator, or a retail analyst can now build an AI company without joining a Y Combinator cohort or hiring a PhD in machine learning. Tools like LangChain, LlamaIndex, and no-code platforms have lowered the barrier to entry so dramatically that the real differentiator is the founder’s understanding of the problem—not their ability to code.

For investors, this is a double-edged sword. On one hand, it expands the pool of investable startups beyond the usual Silicon Valley clique. On the other, it forces VCs to rethink how they evaluate teams. Domain knowledge is now a non-negotiable criterion. Craft’s raise, led by Point72 Ventures and joined by Conviction Partners, reflects this shift. The firm didn’t bet on Levtov despite his lack of coding skills—they bet because of his deep understanding of supply chain inefficiencies.

The cautionary tale? Not every domain expert will succeed in AI. The tools are easier to use, but the models still require rigorous validation. Craft’s success hinges on whether its AI-driven insights actually reduce costs or improve uptime for its customers. Vanity metrics like ‘number of AI features’ won’t cut it anymore. What matters is whether the AI solves a problem that’s costing businesses real money.

One thing is clear: the era of ‘AI builders’ is ending. The era of problem-solvers is here.

Photo: fancycrave1 / Pixabay (https://pixabay.com/photos/hands-ipad-tablet-technology-820272/)