“Language Processing That Accelerates Clinical Research” with Eirini Schlosser of Dyania Health

Nathan (1s):
Welcome back radio entrepreneurs, listeners and fans. I’m producer Nathan Gobes. I’m excited to actually be side by side with Jeffrey Davis this morning. Usually I’m filling in, but today I’m joining him and we’re excited to introduce our next guest Eirini Schlosser president of Dyania health. Welcome Eirini.

Eirini (22s):
Hi, everyone. Nice to meet you. Nice to be here it’s morning.

Jeffrey (26s):
And thank you, Nathan. Thank you for that introduction. Can you tell us about the company and what it does? Exactly

Eirini (34s):
Sure. Absolutely. So we’ve developed a clinical natural language processing. And for many of you that, that may know the industry, or may not know the industry, a lot of healthcare data is buried in physician notes within medical records. And a lot of this data and information is descriptive and format it’s riddled with shorthand abbreviations. And so we’ve developed several layers of our own proprietary tool that can make sense of that information extract, very valuable data points, and then apply that in a de-identified manner for clinical research. So we can match patients, anonymized patients that fit trial criteria for participation, or we can build real world database cohorts for external control, arms and studies, not to go into too much details about what that is, but basically instead of a real living patient cohort, going through a control arm, which would be standard of care plus control that’s when we can do an in silico database profile of a patient.

Eirini (1m 43s):
So a historical profile, a patient that can substitute instead of a real patient, having to go through a standard of care plus placebo in a randomized control arm of the study. So we’re pretty excited about how we’re launching. We work with pharma companies and then we go into and, and partner with the trial sites that are the healthcare systems that they work with for treating these patients that are under clinical studies.

Jeffrey (2m 9s):
So I’m going to try to dumb it down for our listeners who are not medically oriented, what kind of pharmaceutical company would use you and what would be the benefit to them?

Eirini (2m 21s):
Yeah, absolutely. So, you know, we really tried to tackle the most difficult part of the problem first, which were the very, very complex clinical trial criteria that are time sensitive, both for the pharma companies, the patients and the physicians that’s really first and foremost for us, it was oncology. And then secondly, in parallel, we started focusing on autoimmune diseases. So pharma companies that are bringing new drugs to market in those two categories, whether it be from phase one all the way through phase four of clinical studies, we can start with the protocol, deconstruct it and map it out data attributes, and then go find the exact data that would fit their protocol criteria for research.

Eirini (3m 5s):
So they get to market faster instead of having to wait almost in some cases up to two years to find patients that fit trial criteria and in the patients who are seeking access for these drugs right now are found at the time and moment, they need to be found to be able to have those other alternatives after they, in many cases, they’ve tried previous approved standard of care line of therapy, lines of therapy.

Jeffrey (3m 31s):
So this is sort of part of the whole process of challenging what I consider to be the old medical pharmaceutical timeline. And you’re part of that whole process of speeding everything up,

Eirini (3m 42s):
Correct. Exactly. Because normally, you know, a pharma company will call up the physician and say, how many patients do you think you’ll have that all fit these criteria and the physician, you know, it might be a very educated guess, but it’s still a desk at the end of the day. So it’s, you know, a finger near six. And so instead of the physicians giving an educated guess and then having their team do manual chart review and just reading patient files or just waiting for the patient to get referred to them, we can automate that portion of the query down to a pretty accurate degree, as long as the criteria are not subjective or vague.

Jeffrey (4m 21s):
So how did you get into this business? What led, what was the process that led to the starting of this company?

Eirini (4m 28s):
Great question. You know, I want to say I’ve always kind of been very close to the medical field, growing up in a family of all physicians. And I, even as a high school job almost 20 years ago was digitizing my father’s patient record. So that back in the day, those were paper and you had effectively, it was my job to determine who was an active patient, who hadn’t been there in 10 years and then decipher the scribbling on paper. So I’ve always been very close to the space was finishing in biochemistry when I decided to become the only non-physician in my family accidentally fell into investment banking and finance, where I worked on about $60 billion worth of transactions, M and a mergers and acquisitions transactions within a couple of years out of college.

Eirini (5m 18s):
And then towards the end started worked on a couple of technology deals, fell in love with technology, learn to code, did a program at MIT. And then the last 10 years have just been focusing on machine learning. So when digitization of medical records in the U S kind of came to a full, well, I would say partial completion, you know, it was kind of the perfect storm where I had a fantastic network of, of software engineers and teams that we’ve been focused, that we had been focused in the past on machine learning and a great team of clinicians that I could call upon to see basically how does the physician think? And from a data-driven perspective, which is a, it’s quite a unique philosophy because physicians are used to just having everything sitting in their brain and making a decision as they need to on the fly based off of what they have been taught as standard of care.

Eirini (6m 11s):
And so breaking that down mathematically into rule-based methodologies has been a really interesting process overall, you know, I asked the physician, so how many patients do you typically see that would have X, Y, Z characteristic, or what would you, what would be the line of therapy for a patient who has, you know, these characteristics and the answer always starts out with it depends. And so when they give you the answer, it depends. I need to have a mathematical breakdown of almost down to the exact level of like what they would expect to see percentage wise in terms of estimates of how often they see each characteristic on a daily basis.

Eirini (6m 53s):
And so I think breaking down the clinician’s job at that level of something they hadn’t really experienced before, and that’s, what’s been making this the most fun because it’s, you’re right. It’s really been an industry that’s been bogged down by a lot of just manual inefficiencies that they don’t need to be. So our goal was really to help the physicians and help the clinical research process by augmenting it and not to not to replace it.

Jeffrey (7m 19s):
My experience has been, it’s been a very, it’s a traditional industry. They do it the way it worked in the past. If the truck tried and true Now, are you already in a positive revenue model or are you still in the startup stage?

Eirini (7m 37s):
So we are actually, and I think, you know, since we last, since we first spoke, we were just starting our first commercial aviation project. So we are, we are safely into our first pharma contracts and then executing in a few trials sites. So under CPAs with confidentiality agreements with everyone’s, I can’t take, say too much hard details, but we are officially in the revenue phase of the business.

Jeffrey (8m 3s):
Is this your first entrepreneurial venture?

Eirini (8m 6s):
No, this is technically my third. And you can debate a few more ventures that I had done before that as well. So I I’ve been an entrepreneur to the core for most of my life when I first left Morgan Stanley, it was, that was really my first business. And by default, I had to, I realized I had to learn how to code, just to be able to build out something technologically managed teams. So that was really where I got my feet wet. And I would say that between my first business and this business, you know, w what took me two years to figure out then ticks would actually took me about two weeks in this business. So there was a huge learning curve that comes between doing it for the first time.

Eirini (8m 49s):
And I think it’s regardless of age or experiences beforehand, it really it’s different when you’re, you’re building something that’s your own. And, you know, you have the entire risk on your back. And then it’s about building teams that, that can inspire leaders within those teams to take initiative and, and really get things done and drives to drive, to execute for completion.

Jeffrey (9m 15s):
Well, it’s quite a story. And currently you’re located in California if I’m correct, but I know you’ve been in Boston more than I have recently. Where are you Located?

Eirini (9m 28s):
So, actually, I I’d funny enough. I, so I’ve been in California last four or five years have spent most of the last year in hotels, peppered around the east coast. And that’s basically been since the conference and in-person meetings started opening up and, and given that the pharma companies are primarily located around them in the Northeast. And I mean, you can, there are some obviously in, in San Diego and, and, and North Carolina as well, but I think those are kind of our main stomping ground. So I’m usually somewhere in between those, those two places.

Jeffrey (10m 3s):
That’s great. If someone were looking for you or the company wants more information, once they understand that a little bit better, how would they find that?

Eirini (10m 11s):
Yeah, absolutely. So I think you can go firstly to our website@deannahealth.com, that’s D Y a N I a, a health.com and actually we use genius health in, in Greek. So that’s the root of the word, and then feel free to email me at info at, or email our team@infoatdanyahealth.com.

Jeffrey (10m 36s):
Great. I want to thank you for being on the show today. I think Nathan also my producer for showing up, and we hope that as time goes on, you’ll continue to tell us and give us updates on the business.

Eirini (10m 48s):
Absolutely looking forward to it. Thanks so much, Jeff.

Jeffrey (10m 51s):
Thank you. Thank you. And remind everybody. This is radio entrepreneurs.