AI Exchange
+00:00 GMT
  • Home
  • Events
  • Content
  • People
  • Messages
  • Channels
  • Help
Sign In
Sign in or Join the community to continue

Fighting Against COVID-19: How Can AI Help Build & Scale an Effective Pandemic Response?

Posted Oct 06, 2021 | Views 1.2K
# TransformX 2021
# Fireside Chat
Fred Turner
Fred Turner
Fred Turner
Chief Executive Officer and co-founder @ Curative

Fred Turner is the Chief Executive Officer and co-founder of Curative, a healthcare delivery company that has scaled to deliver over 23 million self-collected COVID-19 tests across the US and is focused on creating greater alignment within the healthcare system by vertically integrating health insurance and provider networks. A British scientist from West Yorkshire, Turner attended the University of Oxford. Mr. Turner was named one of the top 100 practicing scientists in the UK by the Science Council 2013. Turner was included in Forbes “30 Under 30” list and ranked first in the European Union Contest for Young Scientist. Turner previously founded and led a16z and YC-backed diagnostics (Dx) startup that built a CLIA lab for validating and launching an STD testing product in Menlo Park, California.

+ Read More

Fred Turner is the Chief Executive Officer and co-founder of Curative, a healthcare delivery company that has scaled to deliver over 23 million self-collected COVID-19 tests across the US and is focused on creating greater alignment within the healthcare system by vertically integrating health insurance and provider networks. A British scientist from West Yorkshire, Turner attended the University of Oxford. Mr. Turner was named one of the top 100 practicing scientists in the UK by the Science Council 2013. Turner was included in Forbes “30 Under 30” list and ranked first in the European Union Contest for Young Scientist. Turner previously founded and led a16z and YC-backed diagnostics (Dx) startup that built a CLIA lab for validating and launching an STD testing product in Menlo Park, California.

+ Read More

Fred Turner, CEO of Curative, sits down with Melisa Tokmak, General Manager of Document AI at Scale, to discuss the challenges inherent in building highly scalable healthcare systems at a critical time — the start of a global pandemic. Together they explore how AI helped Curative quickly scale to deliver over 25 million COVID-19 tests. Fred and Melisa discuss the use of ML to speed up healthcare delivery where existing data silos typically include unstructured or ‘messy’ data. Their discussion includes those critical areas that are traditionally paper-centric, like patient onboarding, insurance, and billing. They consider how best to drive receptiveness with relevant organizations to further invest in AI. How do you scale a coherent end-to-end patient experience from zero-to-millions of patients quickly and efficiently with AI? What tools and processes can you use to scale up teams, infrastructure, and healthcare testing quickly? What are the opportunities for AI to help increase the quality and speed of healthcare services? Join this session to hear the opportunities for AI to enable and scale healthcare to reach more people quickly and conveniently.

+ Read More

Nika Carlson (00:23): Next up, we're thrilled to welcome Fred Turner. Fred Turner is the CEO and co-founder of Curative, a healthcare delivery company that has scaled to deliver over 23 million self collected COVID-19 tests across the U.S. Curative is a leader in on-demand public health service programs and healthcare delivery infrastructure, like mass COVID-19 testing sites and mobile vaccination sites. In early March 2020, Turner realized the urgent need to scale COVID-19 test production in the United States. He led a team of world-leading doctors, scientists, engineers and health industry experts to find a way to deliver mass testing in the midst of a growing pandemic. With a network of over 15,000 sites across over 20 states and three CLIA certified high complexity laboratories, Curative and its managed medical entities have provided over 25 million COVID-19 tests and over 2 million COVID-19 vaccines. Turner is a British scientist who attended the University of Oxford and was named one of the top 100 practicing scientists in the UK by the Science Council 2013. He was included in Forbes' 30 Under 30 lists and ranked first in the European Union Contest for Young Scientists. Fred is joined by Melisa Tokmak, General Manager of Document AI at Scale. Melissa, over to you.

Melisa Tokmak (01:58): Hello everyone. My name is Melissa Tokmak, and I am the GM for Document AI at Scale. Today, I am with Fred Turner, who is the CEO of Curative, and we're so excited to have him. Fred, welcome.

Fred Turner (02:14): Hi, thanks for having me.

Melisa Tokmak (02:15): Of course, we'll start with we want to get to know you a little bit better. So tell us a little bit more about you, and you have a fascinating story around Curative. So I want to hear Curative's origin story after that.

Fred Turner (02:30): Yeah. So Curative is my second startup. First one, which I was doing for about five years before Curative, was working on first agricultural genetics and trying to sequence dairy cows to predict how much milk they were going to make. And then later pivoted to STD testing and trying to target antibiotic resistance in STDs. And so learned how to build out high throughput labs and sort of build out that scale that's required to do testing. Unfortunately, the demand really wasn't there in STDs, but definitely learned a lot that then I think set us up well for Curative when the pandemic hit.

Melisa Tokmak (03:16): That's actually super fascinating. So let's hear a little bit about in the beginning, how did you even get into microbiology and biology through agriculture to the testing and then with the pandemic to Curative.

Fred Turner (03:31): Yeah, so I originally got into sort of biology or biotech when I read Craig Venter's book when I was a teenager and really wanted to sequence my own genome, but couldn't. I was probably about 15 or 16 at the time and couldn't get anybody to let me in a lab and actually play with any of their equipment. Particularly growing up in England, there's a little bit less access to research and things just are a bit more hierarchy driven. People didn't want a kid running around in the lab. So I decided that the best way to actually start sort of running some experiments and doing stuff was, oh, I'll just build the equipment then, and then I can run it at home. So I started off building a PCR machine in my parents' basement and it was kind of yeah, the sort of first foray into things.

Fred Turner (04:24): And I won an engineering competition in the UK with that, and then this, farmer saw it on the local news and wanted to buy the machine to test his cows, but I didn't have, well, I only had one, so I didn't want to sell it. And I wasn't particularly interested in cows at the time. And then he mailed me a box of blood samples with a check taped to the front. And so I thought, "Huh, that's actual money. I'll do these tests." And then he seemed pretty interested in the results. And then his friends started mailing me boxes of cow blood. And it kind of went from there.

Melisa Tokmak (04:59): Pretty amazing.

Fred Turner (05:01): Doing undergrad at Oxford at the time and dropped out after two years to focus on scaling the company.

Melisa Tokmak (05:09): That's pretty crazy. It looks like you just get a check and the sample and go from there. Pivoting has also been a core part of your story. And I find that very interesting because I think it's not easy, it's not easy. And then I think you have to be very thoughtful and intellectual, entrepreneurial to understand that the next thing, and also brave to be able to do that in the moment. And thankfully, luckily for us, especially in Silicon Valley, we see a lot of these and they do end in success in your case as well. Tell me a little bit more about that pivoting part. How did that feel? How did you do that?

Fred Turner (05:58): Yeah, yeah. I think you got to just really sort of dig deep into the data and try and understand the first principles of what you're doing. And then it takes quite a lot once you're comfortable with the idea to then bring the team along and get everybody up to speed with okay, well this is the reasoning. This is why we want to change and go in this direction. Yeah, it's definitely very different pivoting experience at Shield when we went from cows to STD testing, to then at Curative, pivoting from sepsis into COVID testing. Curative happened a lot faster, it was a lot less time to, I think, sort of get everybody comfortable with the rapid change. But we only founded Curative about a month before to work on ending sepsis, mostly just through getting people to actually follow checklists and make sure that you actually do everything for a patient that you're supposed to do based on guidelines and best principles.

Fred Turner (07:03): And the idea was if we can standardize that across all treatment for sepsis, then that could save hundreds of thousands of lives. So that was the initial premise of Curative. And then about a month in, all of our clinical trials got put on hold because the hospitals were all preparing for this pandemic and this was like mid-February. So it wasn't really getting as much media attention yet at that point. But everything that we were doing was on pause. And so we sort of sat down and said, "Well, we know a lot about testing. Maybe we should go and try and help for a couple of weeks, get some testing off the ground and then COVID will go away and we can go back to sepsis." And so initially it was, "You know what? Let's do this for three or four weeks and we'll help out as much as we can. And then the world will go back to normal." Here we are, 20 months later still doing it. And yeah, the team at the beginning of about seven of us at that point, when we pivoted, then grew to a team of 7,000.

Melisa Tokmak (08:09): Oh, wow, wow. That's quite the scale now. And so I have lots of questions about specifically the healthcare part of it and also in microbiology, but before I go into that, something you said is very interesting, which is rapidly scaling. Rapidly scaling and healthcare doesn't fully match for me in my head, which is so interesting.

Melisa Tokmak (08:38): But tell me, and you said that you had to figure out through these ideas that you have been going after how to scale something from ground up. And that started with data really digging into the data to be able to understand basically what's needed next and how things are going. But also when you find that fit in the beginning to how you scale it from there, whether it is building your own labs or equipment or other things, you had to do that very, very quickly and at scale, especially in your latest idea, let alone it's not just the company didn't have time. The world really didn't have time. It had to be quite fast for everybody. So tell us a little bit more about how you got to do that. This is very impressive because when we look at the world, the various companies who were able to scale to that size are not very young, and it did take many, many years. You figured out doing this in a very short period of time. So tell us more.

Fred Turner (09:48): Yeah, it's definitely, I think harder in healthcare than in a lot of other industries. Everything that you do is a patient's experience with the healthcare system. And so there is a lot less room for error than there are in other industries. Obviously, I think COVID opened a lot of opportunities for many companies to scale up very quickly because the need was so great. I think what we've found, kind of the main thing that let us scale it very quickly was vertical integration. And it's something where in healthcare right now, it's a very fragmented industry. And even something simple, like getting a lab test, there's usually like 10 or 15 companies involved in this long chain that the patient has to interact with in some way to actually get the test done. And what we found is when you actually start to bring more of those things in-house, and basically own the entire interaction with the patient, you can really optimize the patient experience and optimize it at scale so that patients get a consistent experience, no matter which Curative site they go to for testing our vaccine anywhere across the country, because we have vertically integrated every aspect of the testing.

Fred Turner (11:06): So we set up the testing sites, we staff the testing sites, all of the software for signing up, making an appointment for software that runs the lab, the software that runs the testing site is all built in house. The labs are run and operated by us, and we build a lot of this stuff that goes into the testing. It's really let us sort of control every aspect of it and optimize every aspect of it. And it's a lot easier to fix problems and iterate very quickly when you're doing that internally than if you have this like group of 10 companies that are sort of trying to iterate together to form a vaguely coherent experience for the patient.

Melisa Tokmak (11:49): Yeah, that is actually different from a lot of the things we're seeing, where sometimes people come in and take really one part of it, one stack and the whole structure and specialize, make that better. But in your case, you're basically not only doing the tech side, which we'll talk to in a second, but also the operation side so you can control every step to be able to do that at a high quality and as fast as possible.

Fred Turner (12:20): Yeah, and I think the integration that you get when you bring all those pieces in house and sort of go full stack in healthcare is you get all the data from every aspect of it. And you can really use that to drive and optimize. And so we have everything runs on one tech stack internally that runs the lab, runs all of the field sites, runs all of the scheduling and appointment software. And so we can really see in very granular detail exactly what's happening and what the trends are and where people are looking for testing, more appointments for testing than there are actually appointments available, where we have too many appointments and we can reconfigure and move things around very quickly. And I think that that's been really important for letting us drive the scale.

Melisa Tokmak (13:08): Yeah. So in a lot of the startups, when they're growing and they're growing fast, infrastructure is something that they think about it, but also the real investment comes in later in the years. But I think you here, building a healthcare infrastructure at such a large scale, and you're working with different states, so many different people is always challenging, but you have been able to handle that very well and able to ramp it up for testing and vaccination programs, not only the tech side, but operations side. And you're saying data and infrastructure played a big role really from the beginning. Let's go into that a little bit more. What do you mean with that? And how did you build these systems from the beginning? Because they were so essential in this case.

Fred Turner (14:00): Yeah, our approach to solving most problems is sort of a mix of pragmatically vertically integrate. So you can't vertically integrate everything on day one, because then you just don't actually start and nothing gets done. So you have to start somewhere, which is in usually a less vertically integrated way, and then basically begin integrating everything that is a pain point. And so we started off with the labs. The labs were the missing component, the very beginning of COVID, there were just not enough labs that could run COVID tests. So we built out the labs and then the bottom layer moved to, well, nobody can get testing supplies, the world's run out of swabs and run out of tubes. So we built out the supply chain piece, and then we basically broke all of the lab software, all of the standard off the shelf lab software just couldn't handle the volume we were pushing through it.

Fred Turner (14:52): And so it would go down for a couple of hours and we'd have 50 people in the lab just sitting there not doing anything. So we started building out bits of the lab software to replace it with an in-house system. And then eventually we built out the field sites, the appointment scheduling, the public health integration. So you've got to kind of do it piece by piece. But I think as soon as we started adding the in-house software component and the tech stack, that's when we really started getting a lot more data on what was going on in the lab. And just very early on built culture of a lot of basically most of the team having access to the raw database and being able to actually just ask questions about what's going on and just using tools like Metabase internally to make sure that a lot of people are asking a lot of different questions all the time about what's happening, how can we optimize? How can we improve things?

Fred Turner (15:48): We also have a daily meeting we call the sync that happens, it used to happen to seven days a week, moved back to five days a week now. There's basically the entire exec team gets on a call and for an hour just goes over all the data of everything that happened in the last 24 hours and makes decisions based on that. And it happens every single day. So when we onboard people, you can very quickly get all of this context of what's happening because everybody is just pouring over all this data every day.

Melisa Tokmak (16:23): Yeah, what kind of decisions would they be?

Fred Turner (16:25): Usually like scaling up or scaling down like lab capacity. What do we expect sample volume to be? Where do we need to open new testing sites? Where do we potentially need to move testing sites? Routing samples between the labs, staffing up vaccination sites and anticipation of new approvals, or it's being approved for a wider age range.

Melisa Tokmak (16:51): Once we make those decisions on a daily basis, since there is a physical world component as well in these decisions, whether it is spinning up new testing sites, staffing, et cetera, how fast does it really take you to be able to implement them?

Fred Turner (17:08): Yeah, obviously it depends on what the decision is. There's a lot of things that we can implement now in software. And obviously that gets rolled out to the testing sites very quickly. So that's been another, I think great piece with the vertical integration is we can make changes very rapidly across all of the sites and we can really tweak the workflow on the site without necessarily having to go retrain thousands of people. We can actually just make that change in software. And that has moved very quickly. Obviously there are real, yeah, physical things like moving a testing site from one location to another that do require a bit more time and somebody has to actually go and physically move it. But I think getting the cadence of decision-making to be very fast, we can then implement those things pretty quickly as well and monitor, okay we started to move these sites from here to here. How's that going? Has volume increased, decreased? What are we seeing? Was this the right decision? Should we change course? So we can very quickly iterate based on that.

Melisa Tokmak (18:15): Yeah. No, thank you so much for sharing that. I think this is good for everyone listening to this here as well, because and we have been lucky, like here and in this country, now we can book testing and vaccinations with basically like five, 10 clicks. But it's so important for people to be able to hear what goes into that by so many people and people like you to make that possible. And it looks like sure, that really has had a wild journey supporting millions of people. So thank you first of all, for everything you and your team have done for the people to make it so easy to get tested and vaccinated. And I think since a lot of your patients, interacting with you and with Curative for these services, they may not realize from the get-go that Curative is a tech company.

Melisa Tokmak (19:09): And we're hearing that here with every effort that goes into that, to be able to make it scalable, at high quality, work fast. And you talk to us a lot about the data, which I'm really interested in it today, because once you have data, you can get really, really awesome things. You can do awesome things with that. So let's go into that. Tell me a little bit more about your teams, especially tech teams. What do they really work on to be able to make these things happen? And how do you use machine learning with all this data that you have collected as well in your day to day to make these processes even more seamless?

Fred Turner (19:54): Yeah, I think we've managed to get an enormous amount done on the tech side with a really lean team, and the team have just done a monumental lift of working night and day to build this entire system from scratch. The whole thing was built in the last 18 months and didn't exist before that. And yeah, we're actively growing that team and really trying to build out this infrastructure for some of the future endeavors beyond COVID has kind of a key piece of healthcare infrastructure, but there is an enormous amount that goes on in the back end to make the experience for the patients seamless. One of the, I think key things that was built out and really had a lot of time and effort from both the tech team and data science team that we have in house is the insurance billing side.

Fred Turner (20:52): And we've kind of been able to set up as from a user standpoint, Curative has never charged a patient for a COVID test or vaccine and never will. So nobody has paid as a patient, any money for getting a test at Curative, and we don't turn anybody away if they don't have insurance or they don't have coverage. We on the back end are able to knit together a bunch of different sources of payment for testing and vaccine so that from the patient's standpoint, they don't do anything. They just show up and get their test or get their vaccine that they need, and we can knit that together on the back end. And so that's where a lot of our use of AI and ML has come in, is working with the insurance industry that is a heavily paper-based, pretty old fashioned industry where we have all this patient data and that needs to get transformed into something that can be submitted to an insurance company.

Fred Turner (21:54): A lot of it is images we have for people that don't have insurance. We're able to bill, there is a federal government fund that covers testing and vaccination services for the uninsured. It requires us to capture a driver's license number. And so we have 4 million or 5 million images of driver's licenses, and we use that to get the information we need to make sure that we can keep paying for the tests, but that it's really easy for the user that they don't have to worry about the fact they don't have insurance. They can still get the testing that they need. And then a lot of the way the insurance industry works is they like to mail back physical pieces of paper, which if you're doing a couple of hundred tests is not much of a problem, but if you're doing millions of tests or our peak testing volume is 200,000 in a single day, then when insurance companies mail you back a single piece of paper or the check attached for each individual test, there's a lot that has to be automated to make that actually feasible.

Melisa Tokmak (22:59): Yeah, no, definitely. I mean, especially first to the point of your team, I personally have been working with you and your team, especially on these processes for patient onboarding and how to be able to do that fast and accurate with super high quality models. And I can tell you now, as you're growing this team, they absolutely have been one of the best engineering teams that I have worked with and extremely easy to work with and you can tell right at every step they're taking, they're looking for how to innovate because they know that a lot of people are counting on them. So as you're looking to grow the team, I'm very, very excited. I can definitely attest to that fact. It is an awesome engineering team to join.

Fred Turner (23:50): Thank you. Yeah, we're growing, trying to grow the team as quickly as we can. And I think yeah, we've built a really high quality, strong engineering team that have made this possible. And we want to keep growing that. If anyone is interested in joining us, you can email Colton, our head of engineering, just or Isaac, our CTO, just

Melisa Tokmak (24:17): Perfect. We'll also make sure that that is on the information so everyone can reach out if they're interested. Switching gears a little bit, let's talk about the broader healthcare industry. You basically spent all your life in healthcare, figuring out different ideas and how to be able to support this. And this has been a field that saw a lot of innovation and investment in the last five years especially. I'm curious to take your opinions about these new waves of companies and advancements and especially how they're changing the healthcare industry in terms of consumer experience.

Fred Turner (24:58): Yeah, there's definitely been, I think a shift towards the consumer engagement. And I think in a lot of places in healthcare, the bar for consumer experience has been so low that you don't have to do that much. Just actively engage with the consumer and not be a crappy experience to really be very beneficial. And so I think there has been, yeah, a really interesting way of the direct to consumer healthcare. And I think so the next piece that's missing is that integration back into the full healthcare ecosystem, where we really need to not have sort of all of these separate apps that exist and just do one tiny specific bit of healthcare, but then don't integrate back into the rest of the system. So if they don't integrate with payers and you're actually just having people pay out of pocket for all these extra services, then it's not available to everybody that should be, it should be available to, it's only available to people who can afford to pay out of pocket for it. And if that data doesn't flow back into the existing systems, then you lose an enormous amount of opportunities to spot trends across people's health data, because you have all these siloed pieces of data about specific bits of their health. And so I think, yeah, the, the sort of next phase of the consumer health care trend is missing all those things back into a coherent ecosystem.

Melisa Tokmak (26:39): Yeah, no, definitely. So okay, that's more about than data that you're seeing a lot of advancements to be able to use that and bring it back to the people so that the experience is increasing. So I've been thinking about this, of course, working with you and working in this industry from the machine learning side. I have been seeing a lot of usage from obviously these companies, including yours on AI heavily to accelerate progress with whether it's computer vision or NLP for testing, diagnosis, document processing and even in surgery and more. So I'm sure you see a lot more than me. What are the interesting applications of AI that you see in healthcare? And can you share some more examples from that?

Fred Turner (27:38): Yeah, I think there's a lot of opportunities in healthcare for processes that are currently run by humans with some specific license where that's always been sort of today, our best way of doing a process. So yeah, I mean pathology, I think is one that has already seen substantial change using AI. And it's getting to the point where you can get the existing regulatory systems comfortable that the AI is probably going to be better than a person at this. And so I think there's sort of two interesting tracks of applications.

Fred Turner (28:20): There's automating stuff like pathology or image analysis. And then there's the more mundane stuff that seems kind of straightforward, but right now is this very data lossy system where we have even just looking at medical records, they're still not really transferable. They're still not in any even vaguely standardized format. And so all of that, the actual useful data that could be driving better healthcare decisions is lost. And just some of these applications of yet knitting together medical records between different providers, data to insurance companies that seem like the more mundane things, but still in healthcare often happen with bits of paper getting mailed backwards and forwards or faxes being sent, actually automating some of that and being able to extract insights from processes that have been largely paper driven. It seems like kind of a more mundane application, but I actually think there's just an enormous amount of missed data and missed opportunity there that could really drive a lot of improvements in people's health.

Fred Turner (29:37): So yeah, some of the exciting stuff, I think tools for scribing primary care or other visits that seem again like pretty simple tools, but once you actually start listening in to millions of primary care appointments and can connect that to outcomes, now you can actually start suggesting opportunities for well, in this, usually when we hear this from a patient, it means X and we would recommend drug Y and we're seeing that you actually haven't prescribed that Y, and just those kinds of little nudges and suggestions, I think have yet room for massively improving healthcare, but seem kind of like mundane tasks, but I think yeah, enormous opportunity there.

Melisa Tokmak (30:31): No, I mean well, AI is supposed to do that, like take away the mundane stuff so people can focus on the parts that can't be done with machines. But I think I definitely agree with you. So pathology has seen quite a bit of use cases along with a lot of the document processing to be able to do this and also the assistant part. How can you have an assistant almost to be able to remind people that what they have prescribed or didn't, what are the notes and if there was something related to the patient's history, and I think it's an interesting junction about that at this time, because I mean in the beginning, you and I, Scale and Curative connected because of that as well. Because everything really in the market and the solutions are more focusing on how to be able to fit these real time, real world data and documents into almost templates to be able to handle.

Melisa Tokmak (31:31): But what we immediately see in healthcare, unfortunately real world does not abide by templates. The data is messy and it will continue to be messy. It is omni-channel. It is coming from really everywhere in every format. So if you try to handle those with anything that some sort of template base, or even one size fits all type of models, it is not getting to the level that it can be helpful. The goal is how do you really build these machine learning solutions and models to be able to be robust to that change and variability. And I mean, I think we have been seeing some great results of course, but it is just the beginning as these applications take off even more so that it's continuously helping the day to day for all the consumer.

Fred Turner (32:30): Yeah. I think healthcare is all about the edge cases. You have so many edge cases to every rule and really to be able to get humans out of these processes, we need to be able to spot the 1% edge case because that's the expectation. That's what a doctor is going to be able to do. And to be able to hand some of these processes off to a computer, we need to have that same level of sort of handling of complex educators. And that's what's missing in a lot of like automation and attempts to digitize healthcare. But I think AI and ML really are very enabling of that and will take us to that next level.

Melisa Tokmak (33:19): Yeah, and hopefully, we'll continue to see that improvement with the Curative products. So before I ask some questions about your past too, like your past in biotech, but a few on the Curative side before we move on to that. So in terms of that automation part, you scale this solution really, really fast, and you're looking for areas to continuously automate to get that efficiency and give a better, faster and more seamless healthcare experience. How have you personally at Curative have been investing in these AI efforts and the automation solutions?

Fred Turner (34:05): Yeah, a lot of I mean we're definitely the early days right now. And a lot of sort of, I think first bits, our groundwork, we've been laying, have been making sure that we're building the right data set and collecting as much data as we can so that we can then layer in the AI solutions on top and we're at the beginning of the journey, I think for actually transitioning those processes over. And yeah, I've been trying to make sure that we're laying the groundwork and collecting as much good data or high-quality data as we can to drive that implementation and then really building out the team to support this.

Melisa Tokmak (34:46): Yeah. And when you look at hopefully one day, the COVID nightmare will be over for all of us, but Curative has been really working on making many other healthcare needs and the infrastructure ready for that. Some of the solutions that you have mentioned, whether your investment in automation, your investment with the relationship with the patients and the providers is enabling that. So tell us, give us a little bit of a sneak peek of these areas, your future investments that are coming and how you are investing on those infrastructure and machine learning efforts now to make that happen in the future.

Fred Turner (35:32): Yeah. I think kind of, I already mentioned our insight from COVID is the vertical integration in healthcare is really what has allowed us to scale and offer the patient experience that we do consistently across the country and sort of post pandemic, what we want to do is take that approach of vertical integration, and sort of take it to its extreme. And the one bit that we're not doing right now and having integrated is the payer side. We still work with other insurance companies, work with the government, and that's kind of still a very messy interface. And I think in order to really drive fundamental change in healthcare, the incentive structure has to change. And so what we're looking to build beyond COVID, we call it internally Curative 2.0, is a fully vertically integrated healthcare system where the incentive structure between the payer and the provider is actually optimized for the patient getting a better outcome and the patient having a better experience.

Fred Turner (36:45): And we think that yeah, the bar for insurance companies in the U.S. in terms of patient experience is incredibly low and that there's a real need for an integrated payer provider solution that actually puts the patient experience first and can optimize for that at scale. And so that's what we're building towards is yeah, full-stack healthcare system. And in terms of laying the groundwork, again it's a lot of data collection right now and making sure that everything that we're doing on the COVID side with the testing and the vaccination is kind of laying the groundwork for building the data sets that we need to optimize going forward beyond COVID and making it easy for patients.

Melisa Tokmak (37:41): Yeah. Quite the vision, of course, not only for Curative, but also vision for the consumer healthcare experience, what the gold standard is going to be. Do you see for the overall healthcare industry, anything beyond this that you see as a long-term vision as well, that Curative may not be tackling right now, but it will be part of it beyond the consumer experience?

Fred Turner (38:10): Yeah. I think kind of a general trend that needs to happen and Curative wants to definitely be a part of this is like reassessing the incentive schemes for preventative care and really actually making the U.S. system is designed to avoid paying for preventative care, which does save some money in the very short term but obviously is not a particularly good strategy in the long term. And I think there's definitely been a trend towards startups providing these kinds of services, but not sort of fully integrated into the existing healthcare system. They're add-ons. They are things that some employers pay for. And I think the trend of moving more in that direction of preventative care and how do you actually make sure you're doing what you can do now to avoid very negative health outcomes later is just something that we have to take on as a country.

Melisa Tokmak (39:13): Yeah. How does your collaboration and relationship with insurance companies and government come into this image to make that possible?

Fred Turner (39:24): Yeah, I think it's taking a more long-term consumer-centric view and sort of working with all of the different parties involved with. It's big payers, it's governments to, I think really lay out that well, if we invested in this now, yes this might reduce the profits of the insurance company in the next 12 months. But over the next five years, it's going to mean that we have a much healthier group of people that are healthier and happier and don't need to end up in the hospital and have really expensive negative health outcomes. And so I think there should be long-term alignment there. It's just setting up the right incentive structure to not sort of think in 12 months blocks. I think a big part of that is the data side and using the data to demonstrate that we actually can make some of these long-term health decisions and we can invest in programs that are not going to pay off in 12 months, but are going to lead to a much healthier, happier population and demonstrating that that can be done using data.

Melisa Tokmak (40:37): Yeah. I mean change in these industries, especially when the big rewards are in the long term is always hard, but I think you and Curative emerged as thought leaders as well that is really talking about this on a data-based manner to be able to prove what the longer-term vision will look like, what are the benefits going to be and how we have to be making the steps to get there sooner than later. So it's quite interesting. It's not only a tech company, a healthcare company, but at the same time have been really carrying this heavy duty of being able to bring different parties onto the table to think about this long-term vision as well.

Fred Turner (41:31): Yeah. I think you have to, for some of these things, where is it a long-term payoff, long-term investment, there's sort of two effects. You have to, I think really get the consumer engaged and people want to have good health outcomes and want to have an easy experience. And if you show them that that's possible, then I think you can use the power of the consumer to sort of drive the back end of the healthcare industry to go in that direction. And then if you can show with real data that you're actually driving improvement and outcomes using this, then the industry is generally pretty receptive to actually doing that. And we've done a lot of work on the testing side with nursing homes and testing nursing home employees regularly to avoid them bringing COVID into the most vulnerable population in a nursing home.

Fred Turner (42:21): And through some of the modeling work that our data science team were able to do, we calculated that the testing program in the state of Florida saved 3,000 lives in nursing homes there over a three-month period. And so by having that data and that capacity to model and calculate this, you can actually show the impact that testing is having. And then that subsequently led to a lot more uptake from the government and coverage of testing and expanding and reinforcing testing programs because you can actually show the impact that it's having. And I think if you can create those sort of feedback loops where you're really demonstrating the impact of some of these preventative measures, that's when you can get engagement and uptake in actually doing it. If you just say, "Well, we need to do this because it will prevent this, but we don't have any data to show why or how." Then I think it's not surprising that there's pushback on that, but if you can really show the improvement and the lives saved, or the improvement in health outcomes that this is having, then I think it's a lot easier to drive change.

Melisa Tokmak (43:32): Yeah, no, definitely. And thanks for telling us all about that. And hopefully we're going to see even faster movement towards that long-term future. But in the meantime, I think we have been all using the products you have been building. So we're very appreciative with all the effort and especially engineering and operational effort that goes into that. And before I let you go, your life prior to Curative was a lot about biotech as well, which is pretty interesting. It's a topic dear to my heart as well. I'd like to be, I'd like to read a lot about it. So before we let you go, I'm curious, especially at this time, we're seeing technologies like AlphaFold that are truly demonstrating that we are at a turning point in leveraging AI for life sciences. But if you talk about this in a broader sense, many people are still skeptical, the value of AI in this domain and how to be able to use it in the right way. So from your experiences there, I'm curious how you see AI being more impactful and how, especially when it comes to using that for life sciences, data becomes so important to build these models.

Fred Turner (45:00): Yeah, yeah. I think we're on the precipice of where the complexity of the biological system is such that you can only get so far with humans looking at it. And so we can kind of scratch the surface with people looking at these things manually or researching them, but really can't get deeper or deeper insights because there's just only so much that a human can take in and comprehend. And I think yeah, some of the most exciting stuff is we can actually take, with large data sets, we can take a lot of these systems much further. AlphaFold is a good example and get much deeper insights into the fundamental biology of what is happening without being constrained by the human mind.

Melisa Tokmak (45:54): I like that. I think-

Fred Turner (45:57): I don't know where it went, where it's going to go next, but I think there's yeah, an enormous number of exciting projects going on at the moment.

Melisa Tokmak (46:08): No, I like that because I mean, throughout this conversation, we talked a lot about the things we can use AI for today, like to immediately improve processes. But also we see a lot that it's used for research, to be able to set in our understanding in these fields so that it can give birth to those applications pretty soon. So it's definitely an exciting future.

Fred Turner (46:32): Yeah, yeah for sure.

Melisa Tokmak (46:36): Fred, of course, in this process, I have been booking COVID tests for myself as well, along with vaccination and everything from my side is quite easy. I go in, and we're very lucky to be able to have this here. We go in, do about five to 10 clicks and really give the information and I get a slot to be able to show up somewhere quite fast. And the whole process has been extremely seamless. Of course, I know that it can't be like that on the back end. You're working very hard and your team is working very hard so that I and others have that experience. And thank you for that. But tell us from the beginning to end, what does it really take for you to do so that I have that experience?

Fred Turner (47:31): Yeah, we really wanted to focus on making it very simple for the users. So like you say, they don't have to worry about all of the work that goes on in the back end to knit everything together and make it work. They can just get a quick, easy test. So yeah, from the I guess very beginning, we opened testing sites. So we'll go into a community, figure out where we think sites would be most easy to access and work with the local community on that. We then hire locally to actually staff up those sites. Some of our sites are kiosks or mobile vans or trailers, and we build those, ship them in, set up the site. And then obviously, on the tech side, I'm actually making an appointment through our scheduling system, try and make that as simple as we can, ask just the questions that we need to ask to get all the information that we need, but make it as easy as possible.

Fred Turner (48:34): And that's one of the things, extracting the information from images of either insurance cards or driver's licenses. It means people don't have to type in all that information themselves. They just take a picture and they're done. Then when you actually go to the testing site again, we're running all of the software that was built in-house for actually running those sites, checking patients in and making sure we have all the information so we can get them checked in quickly. And then the actual logistics of running the test, we have a pretty expansive network on the logistics side to move those samples to one of our three labs, either in San Dimas, California, Austin, Texas, or Washington, DC, everything from chartering flights to move samples directly back to the labs to at one point, we broke UPS in Florida with the number of shipments that were going back and forth.

Fred Turner (49:32): So there's a lot going on in the supply chain and logistics side to actually move those samples quickly back to the lab. And again, that's where a lot of the data and the software piece comes in. It's like planning, okay this many samples were collected in 20 different states. How do we most efficiently get them back to the lab to optimize the turnaround time for the patient? We then run the labs and there's a lot of optimization that goes into how are we staffing? What different shifts are we having? How do we run the flow of samples through the lab to optimize that? Those obviously, then all the patient reporting. And one of the things we wanted to set up that seems really simple, but usually isn't done in healthcare is texting people's results back to them. Usually, or often you have to go into some sort of awful portal and log in, and it's behind a bunch of weird filters to get to it.

Fred Turner (50:26): We wanted it to be really simple for people to get access to the results and make that available over text. And there's a lot in the back end that goes into making that possible and making sure the results definitely go to the right person and not to an unverified number, but it makes it straightforward. And then we do public health reporting as well in all 50 states. So test results go directly to the public health department for contact tracing and setting up all those integrations with different public health departments. Again, it goes on the back end and then if there are any problems, we also have about 150 person customer support team that answers the phones and answers emails, and that's all in-house. So if people have an issue, it's an internal Curative person who will answer the phone and work through solving that.

Melisa Tokmak (51:22): Wow. And it's all within the last two years, two and a half years?

Fred Turner (51:28): Yeah.

Melisa Tokmak (51:28): That is really impressive. It is a giant optimization problem from every step, from being able to set up the testing centers to a consumer, a patient to be able to come to you and schedule, and then getting the samples to the right places, getting the results and running the labs, and then finally sharing it and all the while solving any problems that comes from the patients. So it is a giant optimization problem, but you shared with us today throughout how you from the beginning approach that as a data problem so that you could actually gain those efficiencies quite early on and be creative in the ways that you're solving them. It is truly impressive.

Fred Turner (52:16): Thank you.

Melisa Tokmak (52:17): Thank you so much for joining us today, and I learned so much about what goes into the backend of these experiences that we all have had to go through in the past two years of this crazy pandemic, but thank you so much for sharing everything from how these things came to be very fast and how they scaled. And also more importantly, how you're using the technology and machine learning efforts at every step of the way so they continue to improve. And we're going to see the benefits of those efforts, as you go to new experiences, hopefully as we close out the COVID-19 era and have better consumer experiences in healthcare through Curative.

Fred Turner (53:10): Absolutely. Yeah, thank you for having me.

+ Read More

Watch More

Posted Oct 06, 2021 | Views 1.9K
# TransformX 2021
# Breakout Session
Posted Oct 20, 2022 | Views 201
# TransformX 2022
# Expert Panel
# AI in National Security
# AI Policy & Governance