"I still don't know whether I'm an entrepreneur or not. What I always knew is — give me a problem and I get excited."

Ashutosh Garg spent years inside Google building recommendation systems that predicted what you'd watch next, read next, search next before he asked the obvious question nobody in HR tech had bothered with: what job can this person do next?

That single reframe from who has done the work to who can do the work became Eightfold AI. Two unicorns, 700 people, and a category built from an outsider's ignorance turned into an unfair advantage.

In this episode of SaaS from Scratch, Vani Kola sits down with Ashutosh to trace how a technologist with zero HR domain knowledge built one of the world's most consequential talent intelligence platforms — and what he'd do exactly the same way if he started again tomorrow.

What the episode covers:
— The outsider mindset playbook: when not knowing the industry is the edge and when it will get you killed
— Why 40 out of 50 CHROs getting excited is a large market and 2 out of 50 is a signal to stop, not pivot
— The self-doubt and optimism equation that Ashutosh runs in parallel and why paranoia is a feature, not a bug
— Why talent acquisition and talent management being treated as separate problems was the stagnation nobody questioned until an outsider walked in
— The bias reduction architecture inside Eightfold: skills-first matching, global data assets, and why New York State's audit law is a floor, not a ceiling
— What product-market fit actually feels like not "she said it was a good idea" but "she started expanding the idea herself in the room"
— The 18-month rule: if you haven't found it by then, ask whether you're iterating or just meandering
— Why a founder who won't sell shouldn't start a company, and why that applies on day one and on day one of company three
— Managing leadership transitions without burning the relationship and why the people being transitioned out almost always already know

This is a founder who brought a machine learning lens to a problem that had been solved the same way for thirty years. The insight wasn't radical. The patience to go learn a domain he knew nothing about, while holding onto the one thing he knew better than anyone in the room, is what made the difference. That's what this conversation is about.

‍