Yang Zhao

At Handshake AI, Yang Zhao runs the part of the business that doesn't show up in a product demo. As Chief Business Officer, he owns business and operations development for a venture that quietly reframes what Handshake is for. The company spent a decade as the place college students went to find their first job. Now it is pointing that same network at a stranger customer: the frontier AI labs that need vast amounts of high-quality human expertise to train and evaluate their models.

Zhao's argument is contrarian for someone in artificial intelligence. He thinks the differentiator isn't the tooling. "It's not about the tools and the products and the AI that goes into managing the workforce," he has said. "It truly is access to the workforce - and Handshake has the largest expert network in the US." In a market obsessed with model architecture and benchmark scores, he keeps pointing at the people.

That is the throughline of the work he leads. Handshake AI sells human judgment at scale: subject-matter experts who can write, grade, and stress-test the answers that language models produce. The product is only as good as the network behind it, and the network is what Handshake already had. Zhao's job is to turn a student-recruiting platform into a workforce that AI labs can actually deploy against, then make the operation reliable enough that those labs keep coming back.

He came to the role having already watched this movie once. Before Handshake, Zhao spent four years at Scale AI as Head of Product Deployment and Operations - a front-row seat to the years when "data labeling" went from an afterthought to one of the most valuable functions in the AI supply chain. Deployment and operations is the unglamorous half of that story: the part where a clever idea has to survive contact with thousands of real contributors, messy real data, and customers who want it yesterday.

So when he describes joining Handshake as an "entrepreneurial opportunity to build something new and thoughtful within the human data space," it is not a recruiter's line. He is naming the specific gap he saw - a chance to do the human-data business again, from a different starting position, with a network that was built for people first and AI second.

What makes Zhao worth watching is the order of his priorities. The fashionable move is to lead with the model and treat the workforce as a cost to be minimized. He inverts it. The "thoughtful" in his description is doing real work: it signals a people-first operating philosophy in an industry that has not always been kind to the humans doing the labeling, the grading, and the correcting. Whether that philosophy survives the pressure of scale is the open question - and it is exactly the question his job exists to answer.

He is, by training and temperament, an operator. Not the founder who paints the vision, not the engineer who writes the model, but the person who makes the thing run. That role rarely gets the magazine cover. It is, however, the role that determines whether the vision and the model ever become a business.

Look at the resume sideways and a pattern appears. At Zumper, Zhao led growth and strategy for a consumer-rentals company - a two-sided marketplace whose entire job is matching a person to the right apartment. At Scale AI, he ran deployment and operations for a business that matches human contributors to the data tasks that train AI. At Handshake AI, he is matching expert humans to the labs that need them.

The products could not look more different. The underlying problem is the same one every time: take a large, messy pool of people, understand what each is good for, and route them to where they create the most value - quickly, and at scale. He keeps choosing variations of that problem. The domains change; the puzzle doesn't.

It also explains why Handshake's network, not its software, is the thing he talks about. A marketplace operator knows that the supply side is the hard part. Anyone can build a dashboard. Almost no one has the largest expert network in the country sitting there, already onboarded, already trusting the brand.