Meet Lily Vittayarukskul. At 16, she was interning at NASA’s Jet Propulsion Laboratory and on track to become an aerospace engineer. Then her aunt was diagnosed with terminal cancer. The fallout behind the diagnosis has driven her mission to transform care across the world.

-MV

What’s in store:

Read Time: 6 minutes

WOMEN IN AI 2026

The Woman Who Wants to Change How the World Grows Old

At 16, Lily Vittayarukskul was interning at NASA's Jet Propulsion Laboratory in Pasadena. She had been competing in NASA programmes since she was 12, won a competition that took her to the Kennedy Space Center, and was submitting college coursework while still in high school. Everything was on track for her to achieve her dream of becoming an aerospace engineer.

Then her aunt was diagnosed with terminal colon cancer.

For the next two and a half years, Lily's family cared for her at home. There was the physical daily work, the close proximity of watching someone you love lose herself to illness. And then there was everything that came with it: the arguments that had nowhere to go, the money running out, the family slowly fracturing around a situation nobody had the language to navigate.

All of it paid entirely out of pocket because health insurance is a cruel lottery. Medicare excluded her treatments and, the only safety net available, Medicaid, required spending down assets to qualify. Lily's family was caught in the unfair middle ground: too much to get help, not enough to absorb the cost without breaking.

"We never had a plan," Lily says. "We never had a vocabulary for this topic."

Nobody puts this on a to-do list. Nobody makes a podcast about it. It just arrives.

Without shared language for the decisions they needed to make, arguments replaced conversations and money ran out. That split hasn't been mended. Her extended family has not spoken in over a decade. That's the unspoken damage a badly designed healthcare system can have on families.

This isn't the glamorous startup origin story we're used to. The visionary working out of a garage. The coder in the dorm room. But that heartwrenching chapter is where Lily found the purpose that became Waterlily.

Before Waterlily, she spent years doing product and engineering work in health AI, building a technical background that kept pointing towards this problem without yet naming it.

Her mentor at the time (and one day first investor), Scott Barclay, opened doors to aging experts and worked through the gaps in the industry with her. Then he told her she had a thesis too compelling to hand to someone else, and asked what was stopping her from building the company herself. "It was part courage and part support," she says.

From that point, she ran.

Waterlily is an AI platform that helps families plan for long-term care —the period when someone can no longer manage daily life on their own — before they need it. It's not a medical condition to treat, but a daily reality that health insurance was not designed to cover, and that most people never think about until they are already inside it.

It takes about three minutes: Users fill out a form about their medical history and financial situation, and the algorithm, trained on 500 million data points, predicts what an ageing event is likely to look like and roughly what it will cost, and how much of that care might otherwise fall to a spouse or child.

The platform doesn't tell you what to do. It shows you where you're likely headed, so you can make decisions while there is still time to do so. Over 90% of users say they want to age at home. (Which, by the way, should not be a surprise to anyone who has ever met a human person).

"What that translates to me," Lily says, " is that folks want to age with dignity. They want to be in a place where they have a choice." When people hear about nursing homes, she says, they assume it will never be them, and they stop planning.

The word Lily hears continually from talking to families about their care events is "sudden." When she investigated why health events seem to come out of nowhere, the picture was the same each time. Someone is isolated months before the decline, their calls had stopped, and their personality had shifted before family members registered it.

The research she commissioned confirmed those stories. Loneliness is a strong predictor of how severe a care event will be, and how likely it is to become a dementia event. It gets asked about at intake because it changes the prediction.

When she explains what Waterlily can and can't do, she uses a line her team now teaches: "it's not a crystal ball, but it's a whole lot better than just assuming you're the national average." The risk profile updates each year as health and circumstances change, so the conversation doesn't happen in a hospital corridor after every option has already narrowed.

Lily's ten-year goal is bigger than individual households: Building Waterlily into infrastructure that helps sovereign nations design financially sustainable models of ageing. Every developed country is running the same calculation about who pays for care, and who does the physical work when a generation can no longer manage on its own. She wants Waterlily to be part of how countries build that answer before the crisis is already underway.

The idea has already travelled much further than an early startup experiment. Since launching publicly in 2024, Waterlily has expanded across all 50 states, raised $9.2 million with backing from leading investors such as Insight Partners and companies including Genworth, Nationwide and Edward Jones, and expects its technology to reach hundreds of thousands of families by the end of this year.

Waterlily is free to access at www.waterlily.com. If you want to solve the ageing problem in your country, Lily says, reach out. She means it.

You can get our prompt engineering guide to master AI in 7 days below!

POLL OF THE WEEK

Full results will be at the bottom of tomorrow’s newsletter!

THIS WEEK IN AI

Frenemies, feelings, and a doctor with zero student debt

Meta and Apple put their rivalry on hold to build something together. Anthropic banned cruelty toward Claude because AI might be conscious. And Microsoft built a diagnostic tool that outperforms doctors by a factor of four.

Meanwhile, a startup is selling insurance for when your data centre gets blocked by angry neighbours.

Here’s what mattered:

AI Art

Our Image of the Week

Daily Image Prompt

Submit your artwork to Mindstream →