Xiaowen (David) Zhu

I hold a B.S. in Data Science (AI track) from NYU Shanghai and am currently studying Business Analytics at Columbia.

I care a lot about AI products and about this field itself, and I want to work as an AI agent builder or AI product manager. More than executing against a spec someone else wrote, I want to define the product: taking a real business need from zero to one and actually solving it.

The core of the job, as I see it, is connecting business and engineering. Business teams often can't describe their own need or pain precisely; they can only point in a rough direction. Which steps suit a model, where the model's ceiling is, what output counts as good enough, and how to get the thing shipped and running — all of that falls to product to break apart and push through. That's where I think I add the most value, and it's where most of my last two years has gone: pulling a vague pain point apart layer by layer until it converges into a path a model can execute reliably, then using an evaluation system to check that the path was the right one.

AI is still moving fast — plenty of what was consensus six months ago no longer holds, and nobody knows what comes next. I'm a fairly geeky person, and I'd like to keep exploring in this direction.