Real returns on AI investments.

Kerstin Frailey has spent more than ten years at the intersection of AI research and organizational consequence — leading data science at Meta, designing the first consumer AI product at Charles Schwab, and advising C-suite leaders on what it actually takes to make AI initiatives work. She holds a PhD in Statistics from Cornell and is faculty at Johns Hopkins. Her book on AI and ML for business leaders is forthcoming.

The work

At Meta, Kerstin led data science for supply chain products supporting AI and data center infrastructure. Before that, as first Director of AI Product at Charles Schwab, she designed the company's first consumer AI product and built the internal AI evaluation framework that underpins it. She has also held senior product and data science leadership roles at Asana, Olive AI, Numerator, and GuideStar, and co-founded a startup as CEO.

She built Metis's Executive Programs division from scratch — leading executive education and briefings from EMEA to Singapore and across the country, teaching senior leaders to think rigorously about machine learning before it became a board-level conversation. That work directly shaped how she now advises organizations navigating the same questions at higher stakes.

The research

Her Cornell dissertation used America's COVID-19 reporting as a lens on a problem that turned out to be universal: data quality. She found that California's death counts were revised on more than 17% of reported days — meaning the numbers officials used to make policy decisions were quietly, regularly wrong. Published through Cornell and ProQuest in 2023.

The finding reflects what she sees consistently in practice: data quality remains the single biggest hurdle and the greatest point of differentiation for AI and ML products. Organizations that treat it seriously build systems that work. Most don't — and that gap is where AI initiatives fail.

The approach

Over a decade of leading AI and ML transformations has made one thing clear: change begins with people. That's where Kerstin starts. But it's ROI that sustains that change — and accelerates it.

The knowledge comes from hands-on experience building AI and ML at scale, and it shows. Technical topics become strategic clarity. The people accountable for outcomes leave with the understanding to drive them — and the confidence to make the calls that move organizations forward.

That's what empowers real strategy. That's what generates real returns.

Beyond the work

Kerstin has taught statistics and data science in settings that most practitioners never encounter — including as an instructor with the Cornell Prison Education Program at Auburn Correctional Facility. She has instructed at HER Academy's HERO Code Camp and completed the WMNtech Leadership Program. She was selected for the Pear VC Female Founders Circle in 2023. She currently serves on the board of the Yale Club of San Francisco and as Director at Large for Cornell NorCal.