Pharmacist To Founder:  Why This Young Entrepreneur Built A $4 Million AI Startup To Fix Healthcare - video Dailymotion

Forbes

Before starting her software company, Eunice Wu was a pharmacist. She was frustra­ ted that she spent most of her workdays performing tedious data entry and admin work rather than mixing medicines and advising patients. “It was fueled by my personal pain points working in a pharmacy,” Wu says. In 2023, she launched Asepha with cofounder Can Uncu to let AI software process handwritten prescriptions, verify medical codes and handle paper faxes (still common in pharmacies). Asepha, which is based in Toronto, also offers an AI-powered phone system to handle refills, send pickup notifications and more capably route patient inquiries. Wu says her machine-learning software can cut labor costs by nearly a third within the first month. The startup has raised more than $4 million.    Subscribe to FORBES: https://www.youtube.com/user/Forbes?sub_confirmation=1  Fuel your success with Forbes. Gain unlimited access to premium journalism, including breaking news, groundbreaking in-depth reported stories, daily digests and more. Plus, members get a front-row seat at members-only events with leading thinkers and doers, access to premium video that can help you get ahead, an ad-light experience, early access to select products including NFT drops and more:  https://account.forbes.com/membership/?utm_source=youtube&utm_medium=display&utm_campaign=growth_non-sub_paid_subscribe_ytdescript  Stay Connected Forbes newsletters: https://newsletters.editorial.forbes.com Forbes on Facebook: http://fb.com/forbes Forbes Video on Twitter: http://www.twitter.com/forbes Forbes Video on Instagram: http://instagram.com/forbes More From Forbes:  http://forbes.com  Forbes covers the intersection of entrepreneurship, wealth, technology, business and lifestyle with a focus on people and success.

Transcript

00:00I've always been a very hard worker. I'm a first-generation immigrant over to Canada,

00:03but even then, I never imagined how much further you have to push it in order to be a founder.

00:17Eunice, thanks for joining us today. Thank you for having me. Tell me about what you're building

00:21right now at Acefa. Yeah, with Acefa, essentially a lot of pharmacists are being buried by manual

00:27work. So what Acefa does is we have agents that can automate these manual tasks, such as data entry,

00:33phone calls, prior authorizations, all this paperwork. And each of these agents can essentially

00:38reduce the amount of time by around 50 to 75%. So we give pharmacists back their time so they can

00:44focus on the patients instead of these paperwork. Tell me the pharmacist's life before Acefa and

00:50then after. What is the change you're making? So I actually am a pharmacist. I did my doctorate

00:56at UBC. And it was really fueled by my personal pain points working in a pharmacy, where I

01:02essentially was doing a lot of data entry. I would have to look at reference material. This is not

01:07the stuff that you necessarily need a clinical degree for. And probably 90% of my day was being

01:13wasted on this manual work instead of actually seeing patients, talking to them, giving them

01:17counseling, things that actually require my clinical degree. So where Acefa steps in is we have these

01:23agents that are able to do all this manual work. So instead of having to have a pharmacist involved

01:28at every single stage, they are just involved at the last step. And at the last step, that's when

01:33they are essentially talking to the patient, providing that direct patient care, that direct

01:37connection that really needs that human involvement. So you were a working pharmacist. Tell me how you

01:43went from a pharmacist to a software or an AI founder. Yeah, for sure. So essentially, it was during

01:50my studies and my practice where I became extremely frustrated with the healthcare system. We have an

01:56aging population right now. It'll only get worse and worse. And there's just not enough human capital

02:02to sustain that gap in the future. And that's really where I saw this opportunity where, hey,

02:07we need to turn technology. And so I actually founded Acefa with my co-founder, John. And John,

02:13at the time, was working on the microchip that is powering OpenAI and Azure today. And this was right

02:18when ChatGPT was getting popular around 2023. And he had very similar experiences as well,

02:23where he had many people close to him who struggled to access care in the medical system.

02:28So when we met, really, we were united by our passion to solve this problem.

02:32How did you guys meet? Because you're very different industries.

02:34Yeah, for sure. We met in an accelerator program, actually. It's an accelerator based out of Canada.

02:39Okay.

02:39And we were complete strangers. So it's a program where they essentially take 36 people,

02:45they throw them into a pot. And we started the program as the weakest link. We didn't have a

02:49company. We didn't have business experience. We were quite young as well. But near the end of

02:54the accelerator, we ended up being one of the top companies and now still growing strong today.

02:59Gotcha. And how do you and John work together? Because it sounds like as, you know, he has the

03:03technical chops and then you have the real life, real world problems, pharmacy chops. Like,

03:09tell me how that back and forth goes.

03:10Yeah. We learn a lot from each other. Like, I started learning how to code. I can push things

03:16now. And he learned a lot about the pharmacy system as well. It's very, very collaborative.

03:21I think that's one thing that's really important is when you're looking for a co-founder, it should

03:26be complimentary skill sets, but not just complimentary in terms of the skills where you can

03:31bounce off each other, but actually making up for each other's weaknesses as well. And that's

03:35where we really, really work well together.

03:37Take me to the journey of like, you meet the accelerator, you have this idea, and then how

03:41do you go about starting the company?

03:43Yeah. We, I think we started the company just based off of seeing a lot of issues in the

03:48industry. How we started was actually, I sent out a survey to every pharmacist that I ever

03:53worked with. All the faculty I used to work with, everyone who was at the hospital, the retail

03:58pharmacies. And it was an astounding survey where the results showed that, hey, every pharmacist

04:04that I personally knew at the time really didn't like their job. We essentially went

04:09into the medical system and studying pharmacy because we want to help patients. That's what

04:14really drives us, right? But after you're in the program, you realize that there's so much

04:18paperwork. And this doesn't just apply to pharmacists, but also nurses, physicians. There's just so

04:24much holding you back from practicing to the top of your practice.

04:28Was there one thing you wanted to get rid of first and one thing you wanted to automate when

04:32you started building? Yeah, we started off with data entry, especially because there's so much

04:37data transfer between different portals, different softwares. From there, we expand to a lot of call

04:43agents, texting of patients, being able to help with the patient engagement experience as well.

04:49Tell me about this agent you built because what does it run on? Like tell me about the code,

04:54like what is, who built it and how does it work? Yeah, for sure. We build our models internally.

04:58So with our models, we have a mix of different types of data that it's trained on. Some of it

05:03is synthetic, some of them is real world. Well, the way that we've built it is actually one of

05:08the important parts that I like to keep in mind as a pharmacist is every pharmacist,

05:12every clinician has their own preferences. What that means is there's no one size fits all software.

05:17And that's something that's really important to keep in mind when you're building something for

05:20healthcare. Because clinicians are very particular, if you have the ability to give different

05:27configurations, that that's really something that makes or breaks your software. So we not only

05:32had to learn how to train the model to be very, very high accuracy, but also learn how to train

05:37the model in a way where, hey, it can also adapt different clinician styles. Yeah, this is, you know,

05:41if you're working with pharmaceuticals and pharmacists, this is, these are life and death decisions.

05:46Exactly. How do you, how did you kind of safeguard it, make sure that it was accurate, test it,

05:51and be for the real world? Yeah, for sure. So in terms of our accuracy, that's actually where a lot of

05:56our research goes into, where we are refining our accuracy every day. And today, we're around 96,

06:0197%. Even then, if you have, for example, like a 0.5 error rate, if you think about the breadth of

06:07patients that you're serving, even if you have like 0.5%, that could be hundreds of thousands of

06:12patients per year. So that's why we always keep the clinician in place. We're not meant to replace

06:18the clinician in any way. We essentially have a system where actually a lot of research goes into

06:23the user interface as well, where we have the clinician checking at every stage.

06:27Gotcha.

06:27Then the question becomes, hey, how do we balance the automation with also keeping the human in

06:32the loop as well, without making more work for them?

06:35Gotcha. So it is like a, it's an assistant, it's a sidekick.

06:37Correct.

06:38It's taking, it's in a lot of the faxes and data entry and still the clinicians in charge.

06:42You guys have raised what, $4 million?

06:45Correct.

06:45Tell me, how did you go about that? Like, what are your tips for fundraising?

06:48Yeah, for tips for fundraising, I would say, I was very new to it. If you asked me two years

06:52ago, I didn't know what a VC was. I didn't know what a TAM was. It was a lot of learning

06:57online. And our first round, our pre-seed was actually only $170,000. So we actually had

07:03to stretch that $170,000 over the next three years, but we were very resourceful. In terms

07:08of fundraising itself, I think it's a, it's a very interesting journey. You have to really

07:14be good at storytelling, being able to talk about and emphasize not only your passions,

07:19but essentially where this opportunity to go. What is the future that you see and what

07:24you can create as well?

07:26Very cool. What is the business model? How do you make money?

07:29Yep. We charge per usage. So when we are implemented into the various softwares that the clinicians

07:35use, we make it an important fact that if you're not using the software, we're not going to charge

07:39you. So if you, for example, have spikes in, in patient volume, you'd pay accordingly. If

07:45you have a downturn of patient volume, you pay accordingly. And that's something that's

07:48really important because what we're trying to solve for is, Hey, we don't want you to

07:53overhire when there is a surge of necessity for this type of service. And then you have

07:58to fire people that, that's, that defeats the whole purpose, right?

08:00I see. Very cool. What you, with this money raised, what do you, what's the plan? What is

08:05like the next growth plan for the next year or two for you?

08:07Yeah. I, I, the vision, uh, essentially executing the vision of a CFA and what the vision that

08:13we're trying to work towards is we want to put the entire patient journey into one large

08:18language model context window. What that means is right now, if you go to see a physician,

08:22they have to communicate with a nurse. If you go get a prescription at a pharmacy, they

08:26have to fax your physician. And this causes a lot of delays in care and a lot of issues because

08:31number one, um, you don't get access to your medications. Lots of patients are waiting

08:34up to 40 days to get medications because there's so many approval processes, right?

08:39Um, there's a lot of miscommunication and it's also a lot of pressure on the patient

08:43to be really good storytellers of like, Hey, this is my patient experience. So where we

08:48see this going and what we're building towards is this data fabric where essentially we have

08:52all the patient information in one place. So imagine you as a patient, you don't have to

08:56have to worry, ever have to worry. If you're seeing a new physician, Hey, do they have all

09:00the information necessary to make this clinical judgment? If you're going to a different

09:04pharmacy, Hey, do I have all the information that is needed? Um, and you as a patient have

09:09that comfort. Wow. So you're almost building a whole communication system as well on top

09:12of that. Yes. What is the company look like in terms of cause you're doing, you're dealing

09:16with, you know, obviously you're selling into pharmacies, but then you're also doing a ton

09:20of research and then engineering. What is the composition? The composition is very mixed.

09:25We try to do, um, all everyone on our team is either a clinician or an engineer. We build

09:32this company internally with being pharmacists first and, and being pharmacists first essentially,

09:37that's a really important part because if you're building software for pharmacists, it

09:41should be built by pharmacists who really understand the problems that we're experiencing, how we

09:45would want the ideal software. And that's really how we're, we're communicating with, with our

09:50target audience. Cool. You went from, you know, pharmacist to a founder. What was that journey

09:56like and what was the biggest surprise? It was very rewarding. I will say it's very rewarding.

10:02I think, um, me personally, um, I have this goal of touching 30 million patient lives within

10:08the next year. And that number 30 million is very specific. And the reason being is that

10:14every pharmacist, um, on average, they see around 30,000 patients per year. Okay. And by building

10:20a business, I've had the opportunity to thousand X my learning, I feel. And this opportunity now,

10:27I want to take that and be able to thousand X the impact I can have on these patient lives as well.

10:31What's been the biggest surprise about being a founder?

10:34The amount of sacrifices you have to make, um, amount of hard work. I think I've always been a

10:40very hard worker. I know, like I am a first generation immigrant over to Canada and, um, my parents

10:45worked very, very hard to put, um, make sure that we were like always had food on the table.

10:50And from them, I learned a lot of sacrifice, hard work, how to be resourceful. But even then,

10:56I never imagined how much further you have to push it in order to be a founder.

11:00Wow. What's like, if you were going to give yourself a piece of advice before launching,

11:03what would it be about being a founder?

11:05I think being a founder requires you to be very open-minded. Um, you have to essentially go against

11:14the grain of what you learned in school. And what I mean by that is not necessarily the education

11:19of the content that you learned, but actually unlearning the principles being that, hey,

11:24when you're in school, um, you essentially have one opportunity to submit, uh, an assignment and then

11:29get a grade. And I think after being a founder, that's something that I'm a little bit against in

11:35the, the, the education system, because really you should have the opportunity once you get that grade

11:40to then make improvements and get a higher grade. And that's some, the principle that I think founders

11:44have where it's like, Hey, even if something is not perfect, you should go for it. You should submit

11:48it, get that feedback. And that's the fastest way that you can iterate and learn.

11:51Wow. That's great. What, um, on top, like on, on talking about that, is there a business philosophy

11:57that you prescribe to that's made all the difference?

11:59Yeah. One of, um, the, the people I really look up to is Ray Dalio. Um, the way that he runs his

12:05businesses are a little bit controversial, but one of the principles is, um, radical transparency.

12:10And that's one of the principles that we uphold within our company as well, where, Hey, we need

12:14to be direct with each other. We need to give each other feedback whenever possible. Hey, uh,

12:18the people that I work with, I don't work with them because they're extremely talented. They are.

12:23But one of the main reasons is they're extremely good at challenging me. And, and that's something

12:27that I, I think is really important. Well, when people talk about AI, healthcare is one of the

12:33main targets in terms of there could be a huge revolution, whether it's about treatment,

12:37whether it's about creating new drugs, again, fixing that crazy communication fax machine system.

12:44What are you seeing right now, just in the healthcare field in general? And like,

12:46what is your big prediction over the next five years of AI and healthcare combined?

12:51Yeah. You know, I think healthcare has traditionally been an industry that takes a while to innovate.

12:56Um, like for example, AI scribes actually existed since 2018. Only in 2024 did we see a huge

13:02surge of AI scribes being adopted. I think a pattern, that pattern has started to become broken.

13:07Um, right now we work with quite a few large enterprises where they're actually seeking

13:12an AI strategy, seeking a direction on where can we go with this technology. So I think that we're

13:16going to see a real shift where there's a lot of potential for new innovation to come in.

13:20Great. Eunice, thank you so much for joining us. Thank you for having me.

Pharmacist To Founder: Why This Young Entrepreneur Built A $4 Million AI Startup To Fix Healthcare

Category

Transcript

Be the first to comment

Recommended