At age 12, Lily Vittayarukskul was a rising star. The winner of a national NASA science competition, she was flown to Kennedy Space Center in Florida where she was feted and encouraged to pursue a career with the space agency. Her aerospace dreams came crashing down four years later when her aunt — a “second mother” to her in their close-knit immigrant family — was diagnosed with terminal-stage colon cancer.
Vittayarukskul, her parents and her siblings were plunged into a two-year ordeal of crushing medical bills and the burden of caring for a beloved but seriously ill family member at home. Adding to the difficulty, the stricken aunt had been one of the family’s breadwinners. Family strife also ensued.
“Our extended family very much wasn’t here during this time to support us through that,” she says. “But they had a lot to share about what the best form of care for her would look like. They didn’t lift a finger, and it caused a lot of strain in my family.”
Vittayarukskul dropped her NASA dreams and redirected her energies. “I realized healthcare shouldn’t be this broken, and so I stepped into the healthcare space,” she says. While still in high school, she volunteered at three hospitals, putting in over 400 hours of direct patient care rotations and gaining valuable insights into healthcare.
As a student at the University of California at Berkeley, she studied artificial intelligence and genetics. All this came to fruition three years ago when Vittayarukskul, who is still in her 20s, founded Waterlily Planning and created AI-driven software designed to help families avoid the long-term-care nightmare she experienced.
An AI Breakthrough for LTC Planning
Marketed to financial advisors, the Waterlily software uses 500 million datapoints of demographic and healthcare data to accurately predict individuals’ future LTC needs and costs and presents a range of care options and financial solutions.
Rolled onto the market two months ago, the software is being used by about 100 advisors at a handful of firms, Vittayarukskul says. Although it is now in a “private, early-access stage,” Waterlily plans a “more public commercial launch later this year, probably toward the summer,” Vittayarukskul says.
Prudential Announces Pilot
Waterlily got a major boost on March 8 when John Sieb, chief operating officer of LTC at Prudential, announced that the insurance giant is partnering with Waterlily in a pilot program. Sieb enthusiastically endorsed the Waterlily CEO’s use of AI to solve the thorny problem of long-term-care planning.
“The integration of Waterlily’s AI capabilities into our services could prove to be a game-changer, enabling us to offer unparalleled accuracy, personalized planning, and proactive solutions to our clients,” Sieb wrote in a LinkedIn post.
In a pilot to be launched in a few weeks, Prudential will cover the cost of Waterlily for long-term-care insurance policyholders, Vittayarukskul says. The Helper Bees, a Prudential care navigation partner, will identify policyholders who can be helped by the software, recommend it to them, and provide support as they use Waterlily.
The program will be available to pre-claim policyholders who lack a detailed plan for their long-term care logistics and the costs, and to policyholders who have activated their policies and are receiving long-term care. Vittayarukskul says Waterlily will help the second group predict the remainder of their long-term care journey and make appropriate logistics and financial preparations. In both cases, Waterlily will help policyholders analyze how much of their care their policy will cover, revealing any gaps in their financial planning, she says.
“Waterlily simplifies the process for making an extremely sophisticated long-term care plan,” Vittayarukskul says. “But if a user wants more help with building and analyzing their financial plan, our pilot will include a call to action to allow the user to get connected with a Prudential financial advisor for a detailed planning session using Waterlily together.”
In the works is a second phase where Prudential financial advisors will use Waterlily with clients to build long-term-care plans and explore a variety of funding strategies, including self-funding, insurance policies, annuities and other financial products. “In this case, Waterlily is not selling any financial products but helping the advisor and client analyze different strategies and model ROI scenarios,” Vittayarukskul says.
Although it is now in a “private, early-access stage,” Waterlily plans a “more public commercial launch later this year, probably toward the summer,” Vittayarukskul says.
Deciding Early to Assist Financial Advisors
She explains she created the software because her own experience showed that the healthcare sector was “incredibly reactive” in providing long-term care.
“Once I realized … no one’s solving this for the family, that’s when I founded Waterlily with the core problem being, how can we help families mitigate the devastating effects of caring for aging loved ones? That’s what prompted us to create proactive planning software.”
When she learned that financial advisors discuss long-term-care planning with clients, Vittayarukskul says she decided to tailor her software to them. “That’s what made me step into the industry and dive a lot deeper with these advisors to learn their pain points and how can I better serve them on this topic.”
What she found was that the tools advisors use, including national averages, zip code calculators, and Monte Carlo simulations, are woefully inadequate for charting an individual’s care path. (The accuracy of Waterlily’s predictions is 78%, about 65% more accurate than the industry standard approach, Vittayarukskul says.)
Mining a trove of public and private health data covering almost 50,000 families, she set out to create an artificial intelligence program that could accurately predict an individual’s future health trajectory and long-term-care needs based on their personal and family histories.
At first, she faced skepticism from the advisors she was trying to help. “They were like, ‘Oh, that’s not possible to build. There’s no way you could do that.’ And then when I showed it to them, they said, ‘Wait, this makes so much sense. It’s so cohesive, the storytelling is actually really powerful.’”
Despite the skepticism, Vittayarukskul relied heavily on the input of financial professionals. “I’ve had over 400 conversations with carriers, distributors and advisors to understand not only the workflows and nuances of the industry, but also what was their take as I was realizing the concept of AI for long-term care — the platform itself. In other words, I’ve received a ton of help. This was not a lone effort in any way, whatsoever.”
How the Program Works
The advisor’s first step in using the Waterlily program is having the client fill out a detailed computer questionnaire, which comes in two versions — short, taking about 4 minutes, and longer, which provides more accurate predictions. Most people choose the long version, she says.
In the short version, the client provides their health history and financial information, while the long version includes the health information of their parents. Also included are questions on mental health. “It’s incredibly crucial for accurate predictions,” Vittayarukskul says. “Because we know that loneliness is a very large factor in your health outcomes. So, if they’re being quite honest, answers like ‘I’ve experienced restless sleep, I felt lonely, and I felt depressed,’ are common.”
Client preferences
Next, the client is asked to state care and financial preferences.
For example, do they prioritize maintaining their financial situation, maximizing their care, or something in between? “I see most people answering, ‘I want to balance in between,’” Vittayarukskul says.
Another question asks whether the client wants family members to be actively involved in providing day-to-day care. One answer is a preference for minimal direct care involvement; and at the other end, a desire for heavy family involvement. The answer depends on the culture the client comes from, she says, but most people select minimal direct care involvement from the family.
Once clients complete the questionnaire, the program provides answers within seconds.
The central finding is a projection of the client’s future health and care needs, divided into a series of “phases.” For instance, it predicts how long a client will be able to perform basic self-care tasks like bathing, dressing, walking, eating, and getting in and out of bed. It also predicts when will they need assistance, for how long, and how much.
Coupled with these health predictions is an analysis of the cost of various care options. “We guide you through each of those phases so that you can make sure that this is a care plan you’re looking for or if you want to change something about it,” Vittayarukskul says. “We’ll show you how that affects cost to do that.”
At one extreme, a client could select the scenario where they remain at home, cared for by family members. The program makes clear the human cost of that choice — estimating the hours the spouse or children would have to devote to caring for the client, she says. She offered the example of “Harry,” a hypothetical client who uses the software.
“Harry could look at this and realize, ‘If I need care first, I could totally see my wife stepping in a bit, but I really don’t want my kids to step in for my care.” In that case, Harry likely would avoid placing all care responsibilities on his family, Vittayarukskul says.
The goal is to ensure the client is aware of and prepared for their probable future healthcare needs.
The financial component
Initially, Vittayarukskul says, her concept ended there. But discussions with advisors pushed her to go further in developing the software.
“‘Hey, Lily, so what’s next?’” she says advisors asked her. “‘Aren’t you gonna build out the financial planning aspect of it?’ And I’m like, ‘Wait, don’t you guys already have comprehensive financial planning software? Couldn’t you do this by yourself?’ They said, ‘No, the existing calculators and software’s don’t effectively address, or educate the client on what the different funding options are.’”
So, she added a financial planning tool, showing potential care costs over time and payment options like long-term-care insurance. The software allows the client to compare various insurance plans for cost and coverage.
“Because we don’t sell policies on our platform, we’re an unbiased source,” Vittayarukskul says. For example, the client could do a side-by comparison of three policies, and the program could show that two of them that make no sense for the client on a return-on-investment basis, she explains.
“And we map that unique policies’ actual cost coverage, relative to this client’s unique care trajectories. So even though it says the max benefit is going to be $1 million, we can show you it looks like it’s only going to be $300K, and here’s why.”
Self-Serve in the Works
Although the Waterlily platform is starting to gain traction in the industry, Vittayarukskul is looking to further enhancements and capabilities. For example, the software is currently marketed to financial advisors, but a self-serve version is in the works.
This would allow the client to walk themselves through the results and build out their own care plan. “And potentially where we’re really pushing the envelope is: What if they’re capable of building out their own financial plan themselves to a certain extent?” she says. She adds that they would still need to connect with advisors to get better deals on financial products.
LTC planning Needed Now More Than Ever
Vittayarukskul says the predictions and planning that her platform provides are needed in today’s LTC environment.
“We have this silver tsunami, where there’s 10,000 Baby Boomers who are entering into retirement every single day and more households are having less and less children as a result,” she says.
“A lot of people think. ‘Oh, my daughter will take care of this. My son will take care of this. Something will figure itself out.’” But in reality, children and other family members move away and are not available to help with care, she says.
It Could Have Helped Her Own Family
When asked how her software could have helped her family care for her ailing aunt years ago, Vittayarukskul paused for a moment.
“One of the things that was a massive shock to us — it’s how much we stepped in and how much it financially cost us,” she says. “It came with certain consequences — us having to take a step back from our career, or having to take up another job, or having an effect on our education at that time.”
“And so, by having that level of insight, we might have more strategically planned for what we could financially afford. It may have made sense to have more professional services in the beginning or in the mid middle stage or later stages.”
The software also could have demonstrated to relatives how much time her family was sacrificing in the care of her aunt.
“It caused much strife and broken relationships with us and our extended family. So, for them to see that we would spend thousands of hours of care on our aunt, they could probably see … that’s probably too much for us as immediate family. Even spending five or 10 hours a week and helping us could make a massive difference, as well in terms of overall support and balance.”
In a four-decade career in journalism, Ed Prince has served as an editor with many of New Jersey’s leading newspapers, including the Star-Ledger, Asbury Park Press and Home News Tribune.
