Michael Woo

The bottleneck was never the model

Most founders pitch you a future. Michael Woo prefers to argue about the present. His thesis is almost rude in its simplicity: the large language models are good enough already. The hard part - the part nobody put on a billboard - is the plumbing. The data. The tools. The boring platform that lets a real team point a real AI at a real customer and trust the result.

That conviction has a backstory. Woo joined Scale AI as employee number twenty, early enough to watch a startup become one of the most important data companies in artificial intelligence. He ran the core data product and led a thirty-person team obsessed with operational scalability. He spent years watching what separated AI that demos well from AI that ships - and it was almost always the data and the workflow, not the model weights.

In January 2024 he left to test the idea. With Soham Waychal, a fellow Scale alum who had led engineering at a16z-backed Canal and holds five AI patents, he started Applied Labs. The pitch to customers is unfashionably concrete: an AI agent that answers your support tickets, trained on your knowledge base, that escalates to a human when judgment is genuinely required. A digital employee, not a chatbot.

A year later, in January 2025, the conviction had a price tag. Applied Labs raised a $4.2 million seed round led by Abstract, with Point72 Ventures, Outlander and Tetra joining, bringing total outside funding to $5.2 million. The cap table reads like a who's-who of people who build infrastructure for a living.