He Dropped out of Wharton to Build an AI Healthcare Startup. Craig Limoli’s Story

Hello amazing readers,

Welcome to another edition of PIE, where we interview the brightest entrepreneurs and VCs from Princeton and beyond. This issue’s featured founder is Craig Limoli (Princeton ‘12), CEO of Wellsheet, a company that uses machine learning to improve health records systems. Wellsheet helps physicians manage patient data and serves them the most relevant information when they need it.

Did you see yourself going into entrepreneurship as a kid, or was it a later development?

That’s a great question! I really didn’t seem myself being an entrepreneur until later in my life. Only after working at one of the biggest companies, IBM, did I realize that entrepreneurship is the best way to effect change.

Having said that, I was always kind of drawn to entrepreneurship. My father was a small business owner, and other members of my family were too. I do think that played a role in exposing me to that type of life.

What are some of the problems with the current way that clinicians handle medical records? Can you explain how Wellsheet uses AI to solve these problems?

I really got exposed to these challenges at IBM, as a strategy consultant. I saw how tremendously frustrated physicians were at the level of technology that was being delivered to them, and how even powerful platforms like IBM’s Watson can distract from workflow rather than simplify it.

Electronic Health Record (EHR) systems tend to fall short in a few ways. First, they’re static and unprioritized. A physician seeing a patient’s records in one clinical context is going to see the same thing that another physician will see in a different clinical context. Second, there was a rush to digitize medical records, in large part driven by federal stimulus for adopting such systems, which resulted in systems that prioritize regulatory needs over physicians’ needs.

We’re finally getting to a phase where physicians’ needs are taking a front seat. Wellsheet uses AI to anticipate a particular context (past diagnoses, symptoms, the particular physician, etc.) to predict the set of information that’s most relevant to that one physician. That vastly accelerates workflow and allows physicians to come to conclusions much more quickly. It’s not just your information that’s being used to do this, but all information that we’ve processed previously for millions of similar patient-physician interactions.

How did you transition from working at IBM to working at a startup?

I actually lined up my MBA at Wharton (studying healthcare management) with the direct intention of starting a company to solve some of the problems I was exposed to while at IBM. This allowed me to leverage all of the resources Wharton has to offer. I met my co-founders there, as well as some great advisors and mentors. I spent my first year focused on getting the company off the ground. By the summer of my first year of MBA, we had gotten some compelling traction and funding, and then I actually did not return for a second year. I’m a proud Wharton dropout!

When you were looking for your very first clients, how did you convince them to believe in your product?

It was really a difficult scenario for us early on. There’s this natural “chicken and egg” situation, because it’s difficult to gain the support of a client without an EHR partner, but it’s hard to get a partner without having proved your product with a client. Luckily we had some traction with a small EHR vendor interested in what we’re doing, and got to work with a few clients. Shortly thereafter, we partnered with athenahealth. They have a policy about only working with companies with mutual clients, but they made an exception for us because they were impressed with what we were able to show them. They were able to introduce us to some of their early adopter clients. That kicked off a few pilots for us right away, which led to the initial proof points, which helped us to get where we are today, which is serving the largest health system in New Jersey.

What about making a startup was not how you expected it to be?

I think it was significantly harder to raise capital early on than I had expected. I still remember the naiveté, thinking that I would just brush shoulders with some VCs and they would take interest and put capital in. That’s just not how it works. You need to figure out a way to bootstrap yourself to show them something. There are exceptions, but in most spaces you need to have traction to even be able to come to the table, to have some leverage, for an investor to take an interest.

How do you handle work-life balance when you’re the CEO of a tech startup?

It’s very, very different from any salaried employee position. You’re living and breathing the business, and there’s no way around that. You’re constantly thinking about it. That has its benefits, too, in that you have ultimate flexibility. It’s all about your responsiveness to the business needs. If the needs are such that you need to write a ton of emails a day, and you can do that from the beach, no one will stop you from doing that! I try to spend enough time with family/friends. Sometimes that means I work weird hours, but at the end of the day it’s all about the business results.

—

And that’s it for this issue! Feel free to reach out if you have any questions or comments.

Wellsheet is looking for a part-time business intern for spring 2019. Contact hiring@wellsheet.com for more info. Wellsheet is also participating in the PSIP program, offering internships in business development and software engineering.

Check out our other interviews here!