A look at how artificial intelligence is changing healthcare, and what the future holds.
← All talks
Turing Fest 2020
Right Here, Right Now: AI in Healthcare
Auto-generated transcript - may contain errors. Tap a timestamp to jump the video.
Hi everyone, I'm Stuart Whiting, co founder and CTO of Current Health, a fast growing health tech startup headquartered here in Edinburgh. Clearly, healthcare has been at the forefront of twenty twenty in many ways. But today I'm going to talk a bit about a huge area of growth and impact right now, that's AI in healthcare.
I'm going to start with a quick intro of who Current Health are, what we do, and how we use elements of AI. From that, I'd then like to speak a bit more broadly about some of the challenges in healthcare, both before twenty twenty and indeed things that have happened this year that I think AI may really be quite helpful.
It's never going be a silver bullet, and it doesn't provide the solutions to a lot of serious problems that are happening right now in healthcare, but there's certainly some big opportunities here. I'm going to carry on and talk through some of the buzzwords and define what AI really is.
It's had a lot of marketing, and the reality is it's really quite simple. You just need to understand what's going on under the hood, and really, it's all about the application of how AI can be used successfully. I'd like to talk a little bit about where AI is being used already, and more than anything, give you some ideas on how you can bring successful AI products to healthcare as well.
So firstly, I'd like to thank the Turing Festival organizers for really going ahead against the odds and giving the opportunity to talk here. While health tech is clearly surging forward this year, it's really been quite an unrelenting year for startup communities around the world.
More than anything, we need to regroup now, and it's clearly time to build and rebuild like never before. At the same time, really want to point out that what's happening right now is shaking formidable institutions, and it's got to be said, it's really opening some once in a lifetime opportunities right now.
Entrepreneurs willing to begin their journeys right now are going to have some really huge, huge markets to attack. Although Current Health was officially founded in twenty fifteen as a company, a lot of the foundational direction and ideas that led to this being possible it really started out in the financial crash of two thousand and nine, and so out of these situations, good things can happen, and I really am very, very hopeful that we will see a new generation of great startups emerging from this chaos.
So first, a quick bit of who we are at Current Health. Now, our mission is really quite simple. We are transitioning healthcare from hospital to home. We really fundamentally believe that people should not have to go to hospital unless there is absolutely no other option.
Experienced this, both me and Chris when we first founded this company, we experienced what it was like to have a family member bounce back and forth to hospital and go through the pain and suffering of this, and so we realised we had to do something about this, and that was really the founding premise of what this company is about.
When a patient goes home, for all intents and purposes, healthcare professionals have no way of knowing what's going on with that patient. They need to keep an eye on them, they need to identify early risks and deal with those sooner. If they can do that and spot those patients that are at highest risk, then ultimately they can stop those deteriorations and treat those patients while they're still in their home.
Good for the patients, it's good for the healthcare professionals. The patients have better outcomes, and hospitals have less unexpected admissions, readmissions, and reduce those length of stay. So there's a lot that can be done with this, and really at Current Health, what we're providing is that risk management platform to keep an eye on those patients, whether they're in or out of the hospital.
We founded Current Health, actually, as a company called Snap40, exactly the same company, we just changed names last year, because Snap40 really wasn't a very good name for a health tech startup. As I said, followed personal experiences, really experiencing the problem we're trying to solve here.
We are now about seventy people across the UK and US, and I think it's fair to say as of this year in twenty twenty, we're now experiencing hyper growth unlike anything we've experienced before. Very shortly, we'll be monitoring patients in about ten countries across four continents.
In total, overall, we've raised about twenty five million dollars in investment and are now looking towards a Series B round, so keep an eye on that for news on that coming soon. We focus really on working with customers both in day to day clinical care, so that's your typical home healthcare hospitals, where those are typically quite unwell for our patients out at home, and we work with those customers to make sure those patients are well looked after and safe in their own home.
But alongside that, we also do a huge amount of work with biopharma, so things like clinical trials, how can we scale up large scale clinical trials, testing of new drugs and therapies across a range of spaces, and indeed around the world. Our work means that we work with a lot of different interesting organizations.
In the UK, that means the NHS, and lots of different hospitals and demographics and patient populations and diseases. In the US, we work a lot with major health systems, and a lot of those biopharma customers that we work with are working across everything from oncology to respiratory cardiac therapies.
Indeed, right now, we're working with some of the world's most important clinical trials to allow them to move fast and safely. And it's fair to say the last six months have really been extremely intense for us. Startups can be chaotic at the best of times, but this has been beyond anything that we've certainly experienced in the last five years of our journey.
I have to say, our team has really been absolutely incredible the whole time. Everyone in, certainly everyone both in our team and in our contracts and everyone around us has really lent into this crisis as it's been unfolding. We've been really moving mountains, I have to say our team has been truly spectacular at that.
But thinking more broadly at healthcare, you know, everywhere right now, it's really very clear that healthcare has got a lot of problems, you know, these predated COVID. And really, has just kind of exacerbated the whole situation. And healthcare is a really challenging domain to build tech in.
And actually, we obviously are clearly in a time where there's lots of opportunities, and there's lots of chances to have impact. But AI is never going to be a silver bullet in any of this. But it may at least offer some part solutions to some of these quite hard problems.
To name just a few, aging population is very clear that people are living longer, they're living with more complex conditions, and that's really adding a lot of strain to healthcare systems that just haven't got capacity to deal with this. So we're having to think, how can we manage patients outside of a hospital?
Should we be building hospital beds? Probably no, we really should be looking at ways to take the less acute patients, the patients don't really necessarily need to be in hospital, get them back home and keep them at home for longer, that's really what patients really want as well.
We know what's happening right now is causing huge waiting lists, and we know people are not being screened for things like cancer. This is going be a disaster waiting to happen. Long term, there's going to be a lot of problems arising from this in the coming years that are going to be crises one after the other, and we know this is coming.
So, not least right now, we're dealing with some serious issues around staff safety. Clearly with an infectious disease, it means that staff doing what were previously really quite straightforward, simple tasks has suddenly become increasingly risky. Take, for example, checking someone's blood glucose. In an intensive care unit, you would do this every hour continuously, and you would send one person to take the blood glucose, and one person to go and then, say, treat with insulin if needed following that.
That's a problem when you're considering that every single one of those people going to see that patient is going need a set of PPE. We need ways to automate that, we need ways to improve that. Irrespective of where we are now, we need to stop this happening in the future, that means better ways of understanding public health and predicting, you know, and both detecting and predicting ways of managing these types of situations.
And part of that's about how do we get better drugs to market quicker and safer, and that's clearly a huge area of concern right now, that getting drugs to market takes ten years and costs billions of dollars. That's a problem when we're trying to move quickly, we have to find better ways of doing that, but while still maintaining the safety.
Technology clearly has some opportunities there. Underscoring all of healthcare right now, ongoing evergreen problem is that we're spending more and getting less back for it, and that's a problem that we have to deal with if we're to keep on scaling health care and making it work for everybody, is that it can't be new, it can't cost cost, you know, huge amounts of money.
So I'm gonna quickly pause for a moment. And I've kind of been speaking about AI as part of that machine learning for the last few minutes, but there's a bit of an elephant in the room of what exactly is AI, or artificial intelligence, and I suspect you're all probably very familiar with a picture of a robot you know, touching the head and lights going on, and like neural brains and all kinds of stuff, thinking lights, this kind of connotation of Terminator and, you know, genuine intelligence and cognitive reasoning,
and it's really just all rubbish. It really doesn't help the whole cause of what AI is, it just causes fear and concern. At the end of the day, it's just simply cognitive automation. We're just trying to repeat the human thought process, really in a quite a naive and dumb way, and doing that in a, I suppose, trying to replicate some of those more complex processes and repeatedly run those again and again.
And that's really actually quite a straightforward, simple concept. It's just not so not so good for marketing, I guess. So, you know, with that in mind, what what really is AI in this sense? Well, you know, let's keep it really simple. AI is you scoop up some data, usually that data has got a ton of problems with it that you either don't realize or soon gonna realize, you're going to write some lines of code and probably take some some open source software that pretty much does all this for you,
and you're just going to wire this all together in a relatively straightforward piece of code, just need to understand what's going on under the hood, and, then get some magical predictions out of that, that are probably going to be wrong, and to varying degrees of right or wrong.
And, you know, it's really not that much more complicated than that. Again, the real value here is not the AI, the model, the algorithm, you know, it's really about how do you make any of this work in a day to day environment as intensive as healthcare.
As you can see, when it comes to predictions, when they're wrong, they tend to be ridiculously wrong, safe AI is all about understanding why, and then how do you avoid that, and that's particularly important, obviously, in healthcare. AI is clearly nascent tech, everyone's talking about it, it's the big hot kind of cool thing right now, but AI as technology is a concept that's been around for thirty, forty, fifty years, even kind of the latest deep learning AI ML advances are all coming out of concepts that were written in papers forty years ago.
What changed is we now have huge amounts of data, compute power, the cloud, ready to go and deploy and solve these vastly computational problems, that's opening up some cool opportunities, but it doesn't fundamentally solve the fact that healthcare needs to be ready for this type of technology, we need to figure out how to make it successful.
So what do we mean by learning machines? Well, you're probably very familiar with the notion of writing a piece of code, a program. You take a specification where you wanna encode some reasoning into a piece of code to take some input and produce an output, and you specify those rules, you get some test data, you write your piece of code, and then out pop your answers.
If it's wrong, you go back and fix the code, and you fix what's in the middle. Pretty straightforward programming, really the standard way of doing things, and that makes a lot of sense when those rules, that specification are quite simple and well understood.
Learning and certainly machine learning, and by extension artificial intelligence, which extends some of these concepts, it's really about turning this on its head and saying, well actually, in the world at large, we actually can gather answers, we call these labels, from some question that gets posed in the world, and we can go and get lots of that data, for instance, diagnoses of people that have, say, instance, got heart disease, we can get data about those people, demographic, maybe some of their clinical parameters, and we can go through all that into a magical computer program
itself that basically goes back and figures out what connects those labels to that input data, and we therefore essentially get a program to write that specification for us, and that's really all that machine learning is. And AI extends that idea really in as much as that when it comes to data, we build deeper representations of what that data really is, and how do we represent it, and AI extends that learning process of how do we encapsulate a problem, that's why AI is unique, but also inherently risky,
because it takes away some of the human supervision for some of that process, but sometimes that does make sense. So thinking, how can AI actually help us in terms of different areas of problems that we can focus on here? Well, there are really three areas that I personally feel that AI's got potential, huge potential right now in healthcare.
One is really those high volume, well defined, simple, repetitive tasks. These are the things we really know how to do well, and for the most part, eighty nine percent of the time, they're straightforward, there's no complication, and really quite simple, it's about consistency and frequency, and these eat a lot of time, and it's the kind of thing people spend a lot of time doing, but really, there's no value in it, except when something goes wrong.
And so these are things like taking vital signs, clearly an area that we're working on as part of remote monitoring. We need an algorithm to keep an eye on that data, that volume of data all the time, because we couldn't expect a human to do that, so an algorithm does that quite nicely for us, and we can very narrowly define what the algorithm does and the parameters in which it does it, and if something's not right, we can pass that off to a human and get them to deal with it,
and the reality is then we massively reduce the amount of work being sent to the human, while still having a human in the loop for safety. We can also, with algorithms, do a lot of things around assisting labor intensive, cognitively demanding tasks, so things like, how do you pass all this information in a radiology report?
How do you look at all these images and try and make some sense of what's going on with respect to the patient population, and past background conditions, and some of the information that's contained visually within those representations, the different x rays and MRIs, and that's really quite a complex problem, and an algorithm isn't going to be able to capture all that rapidly, but are actually a lot of conditions that are very similar, and a lot of other pathology, probably quite similar, you see it time and time again,
and an algorithm there can actually act as quite a good safety loop to augment human ability, and I think there's certainly a big opportunity there. But things get really interesting is where we can go beyond what we do right now as humans, and that's these new areas of research and exploration, where we can find signals that right now no one's using them because they're just so intrinsically complex to model and understand, and far beyond human cognition in terms of the volume of data and the nuance of that data to be able to make sense of.
So literally just a few days ago, a publication was made, which looks at blood vessels in the eye, modeling those blood vessels, and the width characteristics of those in lots of different people, and it turns out you can learn a lot about someone's cardiovascular status from the characteristics of the blood vessels in the eye.
Makes perfect sense, but there's no way we could do this as humans at any scale, but an algorithm, training an algorithm to do this, incredibly yields all kinds of insight, and these are the kind of findings that potentially could change medicine for the years to come.
I suspect many of you are very familiar with seeing the peak of inflated expectations and the age old, suppose, trajectory of new technologies, how we go through all this kind of grand ideas, everyone realizes it's a bit rubbish, and then out the ash, suddenly we start figuring out how to solve the real problems.
I suppose really, what I'm looking at here is, these are the real problems that we need to deal with, it's not the algorithms, that's all getting commoditized, and you pretty much take it off the shelf and apply as you want now. People, stakeholders, healthcare is complex, because there's so many people, so many stakeholders involved.
Who's paying for it, who the patient is, who the family are, the expectations of healthcare professionals, all of these all interact in ways that change all the time. And building any product in this space is going to require you to really understand the people that will be involved and how they are going to act upon what you're doing.
And you're going have to figure out how to deal with their concerns and issues. And that's hard. You know, it's hard in any product, particularly hard in a product where the stakes are so high. Building AI is great, you know, you can throw a piece of data at an algorithm, and great magically, you've got some prediction, doesn't mean it's clinically very useful.
And that's a real common problem in AI machine learning, certainly, universities build incredible algorithms, publish papers, but go and ask a clinician, and they look at it, like, really doesn't have any real value, because by the time you can give us that prediction, we knew that this event's already going to happen.
Trust is a big issue, clearly, people just don't trust this technology right now, and we've got a heck of a journey to try and change that, and that's as much about data security privacy as it is about understanding what AI is and isn't and how we make it successful and frame what AI is suggesting as a suggestion and highlight, know, explainability and how we bring all that together into something that really kind of lowers the risk of this being valuable in healthcare.
Evidence does, does this provide any real value anyway, no one's going to buy a product if there's no evidence. And the problem is, you can't get evidence unless people are using it. And the problem is, you know, people using it takes a long time to reach the bar that is expected in healthcare.
And that makes building products in healthcare really challenging and quite incompatible with small companies. Healthcare is a system as well, like dealing with the workflow of how people do healthcare, how people provide healthcare is difficult, and the risks that emerge from that as well.
Safety, regulation, if you're building an algorithm that makes a suggestion on someone's health, or some recommendation, clinical recommendation, even so much as telling them to go and see a doctor about something else, there's a good chance you're building software as a medical device, it is a medical device, and regulators, you know, will come down heavily on you if you fail to meet your obligations there.
Above all else, economics as well. There's a very complex chain of economics here. The NHS simplifies it, because I suppose really, have a universal health system that ultimately kind of pays for everyone's healthcare for the most part. Go and operate somewhere like the US, and all of a sudden, there's sixteen different people satisfy that there's some kind of return on investment on your approach.
So what does the road ahead look like? Well, tough, I think is probably one word that pretty much describes it. Moving fast in this field, it's clearly a space that can thrive, I think here, there's great opportunities, but, you know, let's face it, moving fast in healthcare, you know, can be dangerous, and you're always going to have to meet the safety obligations.
And, you know, that's not always compatible with small companies moving fast. And that's where a lot of people have gone wrong. I'm sure a lot of you heard about Theranos and some of some of what happened there. These are the kind of things that, you know, really scare a lot of people that are genuinely working in health tech and trying to do the right thing is that's a place you just don't want to be.
Regulators are getting to know, you know, the space more, and they are encouraging and adopting, you know, new ways of doing things. But at end of the day, their job is to keep patients safe. And we have to make sure that we work with them to figure out how we do that, at the same time as bringing innovation to market that can improve things.
Data, clearly getting hold of data is a big problem, you know, privacy security concerns aside, if you haven't got data, it's gonna be really hard to do much in this space. You know, that adds all kinds of difficulties on to, I suppose, really kind of the early starting startups that just simply don't have any data and getting things at scale to have any big impact is going to require some interesting strategies.
The tech here is, you know, for the most part is pretty straightforward. It's come of age, it's really straightforward and simple. You can pretty much get what you want off the shelf now, the tech is not going to be a competitive advantage. Biases, we've spoken a lot about already, it's pretty clear.
The GCSE fiasco of, indeed, all exam fiasco this year of predicted grades have just shown biases in data, when you take data, you know, without understanding structurally, you know, why, how, what is cause and effect in that data, you're going to measure things you don't realize you're measuring, and if you're making predictions, those predictions are going to wrap all those up in rubbish.
And the chances are you're going to cause issues, and we know that all too well from this year. And the trust obviously is a core part of that. So, you know, there's big opportunities here, you know, overall, but probably the biggest change that's coming our way is that, you know, in healthcare, there's a generational change.
People that are developing healthcare now into the future are people that have grown up with this type of technology, so they want to adopt it, they want to embrace it. So now more than ever, we've got people that want this type of technology to make it into healthcare and be successful.
So actually, we've got some big insiders, but, you know, let's face it, we're going to have to focus on the lower stakes, kind of lower risk, easier ways of getting healthcare AI up and running, magical algorithms that kind of diagnose and, and, you know, satisfy some of those high risk use cases are just not going to happen for a long way off yet.
For the same reason that we don't trust autonomous vehicles to drive around without a human at the wheel, we only trust them to go and park themselves in the garage. You know, parking in a garage is a low risk endeavor. It's low stakes, driving down a motorway at seventy miles per hour clearly has very different consequences if it goes wrong.
You know, until we figure out things like privacy, risk management, explainability, all those kinds of factors, there's just no way that we can deploy AI at true scale in healthcare. But that's not to say we can't start solving some of these initial use cases.
And I suppose that really brings me to one final thought here. Hear this quite a lot, and it always frustrates me, because this never comes from anyone that's ever worked anywhere in healthcare, and it's this statement that AI is going to replace doctors.
And I think it's fair to say that AI will just simply never replace doctors. Medicine, healthcare is an innately human endeavor. There's just no way that AI can ever replace what humans provide in healthcare. It's just not going to happen. But that said, I think it's fair to say that, you know, doctors using AI as a tool, you know, and that's really what AI is, it is a tool, are probably going to replace doctors who don't.
And I think there's some huge opportunities for startups to come move into this space and start doing great high impact things. Excellent. Great stuff. Stuart, thanks very much. So we just heard a little bit about AI and health care from Stuart. And we're going to bring him back in for now for some Q and A.
Anyone out there who has any questions, feel free to ping them across in the event app. I'll have a chat with Stuart about them. But just bringing you back in, Stuart. So thanks for all of that. Thank you. Pretty incredible space that you guys are operating in.
It's not like you're making some martech app or something. You're actually building something that might change, really change people's lives in a pretty significant way, and change even systems. I'm really interested to your comment on health care as a system. We'll maybe dig into that.
But I just wanted to start with something I've been doing this year. So I've been watching, rewatching for like the millionth time Star Trek The Next Generation quite a lot. I think it's like Captain Picard. We need a bit more Captain Picard in twenty twenty given the leaders that we have around us.
And we've got Doctor. Crusher in Sick Bay on the Enterprise. We've got Tricorders. We've got all sorts of amazing stuff. Okay, it's four hundred years in the future, but, I've also this year, my son was in hospital for some surgery and, which all was excellent and all great, but the consultant was trying to show me some scan images on his desktop in his office.
And this guy is like an incredibly talented neurosurgeon. And he's operating with Windows XP, just like trying to get access to these files. The world that you're talking about, which is sort of heading off towards Doctor. Crusher, and the world that we're dealing with right now where Windows XP is still hampering people, how do we go from one to the other?
Very, very good question. So the realities of healthcare right now is that we haven't got the basics in place to do any of this anyway, like the AI piece is kind of the big picture, years down the line, you know, we can start kind of developing, I suppose isolated solutions.
But the one word that everyone talks about in healthcare right now and healthcare it is interoperability, that we have all these bits of data flying around all over the place, some are still written on on bits of paper that facts through, we have to figure out to get all of that into one place to consolidate that into medical records that actually contain all this information before we could really do anything good with it.
And we've got a huge journey to make that happen. And, and, you know, as you said, your Windows XP like that, that's part of a function of a healthcare IT infrastructure that's really kind of, you know, that's come from the past and kind of serves a use case now, but sure as hell isn't gonna serve what we need to go, you know, what we need to do to bring AI to scale.
So you're absolutely right, we've got to fix some of that plumbing and infrastructure. What about at the other end of the scale? Because whilst, you know, we maybe see funding challenges in healthcare systems resulting in, I'm assuming that when you go into surgery, the tech that they're using in the theater is better than the Windows XP on that computer.
But at the other end of the scale, we're seeing consumers increasingly all buying more tech than we ever have. And Apple just came out with Apple Watch six. And we've got echocardiograms. We've got blood oxygen or sort of blood oxygen indicators, it seems to be a bit unclear.
What's the direction of travel for consumer health tech? How far removed is that from your world? And how real is that as something that can really help our health? Absolutely. So there are two, I suppose two kind of things to point on there.
One is the you mentioned about, you expect, I suppose, technology and sophistication reliability to be be pretty good in in healthcare itself in operating theater, for instance. Actually, that's turned out not to be the case in many cases, if you go and take a look at some of the FDA recalls, some of the medical device recalls, you'd be quite scared about the lack of security in particular.
A lot of things like infusion pumps are operating without even the most basic security. And that's quite scary when you think actually, you know, yes, they work well as a medical device, but they really actually cut corners on what we consider consumer grade security, which is something that the industry has to deal with.
But I suppose kind of going on from that, know, what does that mean for consumer and consumers obviously driving a lot of this space forwards right now, this kind of notion of an old school bedside monitor that sits there beside a patient in the hospital and a patient has to go to hospital to have any kind of notion of monitoring is clearly getting eroded now with with good consumer tech and consumer tech is forming what people expect the healthcare experience to be end to end, not just the device,
but right through to the care, the counselling, the guidance that, you know, that human in the loop aspect there as well. And certainly with Apple, and these guys bringing these devices to market, fantastic for really, I suppose that their main focus is really around wellness.
What they don't actually have is a medical device that can be used for clinical decision making. And Apple, you know, and indeed, a lot of the other companies work in the space are quite careful about how they frame that it's, it's a number that can be used to kind of give you an idea of your health.
But a doctor wouldn't be able to use any of that in any kind of serious decision making situation. There's a secondary question there whether or not they should be able to and I personally feel that you know, healthcare and consumerization to realize scale of good healthcare is a consumer problem.
We are all consumers of healthcare. And actually, we have to fix that. It's not eye for all. It's about how can we bring medical grades, monitoring medical grades, techniques into day to day life and do that in a meaningful way. And an Apple is kind of coming out from straight consumer, we're going at it from kind of wearables on clinical, I dare say, we will probably meet in the in the middle in the future.
And I guess we got to figure out who's going be the big companies in that space. So on that kind of segueing in, Katie Beaton has sent in a question asking about what solutions are being requested from health care providers or patients, which I guess is the foreshadowing of where consumer and professional health care meet?
Yeah, I think that's actually a really good, interesting kind of question right now is that, I suppose, for a lot of patients, certainly our generation is the idea of kind of signing up to go see a GP and then having to turn up a place, you know, a GP practice and wait for twenty minutes in a waiting room and then have a ten minute appointment.
And then and then go home and then you know, be told that we'll actually come back in two weeks, see how it goes in the same again, taking time off work each time, like clearly is not great. But that said, a doctor doesn't really have the tools to make good decisions or the data to make good decisions until you come into that room.
And so we have this, you know, two sided issue of, well, doctors can only make decision what's in front of them. But simultaneously, you know, unless we're physically present, we can't get that to them. So you know, what we're seeing is a lot of healthcare providers trying to figure out how to collect that data outside outside of the clinic.
And that's great. And likewise, you know, if patients are sick, are genuinely sick, then get into hospital soon to deal with that. You know, issue is that we see healthcare from our own perspective. And you know, we're young, fit and healthy. So for us, the idea of capturing our own data at home and sending that to a doctor and then turn our say, it's fine.
It's all good. Don't worry, it is easy. The real issue to healthcare right now is the the elderly frail patients who have never used a smartphone before. And these are the ones that you know, really are the most costly in healthcare right now.
And the ones that are obviously shielding from, from COVID. And really the ones that are driving a lot of the problems that need to be dealt with, because these are the ones that that when they get sick, you know, things go horribly wrong.
And solutions like we know them in consumer healthcare right now just don't work for these patients. Yeah, interesting. So there's, there's going to be a lag, I guess, partly because of the let's call them end user. It seems a little inhumane, but versus people who are used to tracking, monitoring.
We all maybe use some sort of app for maybe it's Strava for running or whatever it might be Fitbit and all that kind of thing and then I've actually got my mum to start using Strava recently I was happy about that Brilliant! Yeah and she's looking at it and she's like oh you know my parents are obsessed with counting steps and all this kind of stuff I guess there is you know we are even the older people in society are moving that way.
You mentioned something about healthcare as a system which I touched on and this question might be I guess this probably touches on some of the frustrations and challenges. How much of the challenges for you and for what you're trying to do, how much of those challenges are human versus technical?
I mean there's always the inevitable technical challenge of trying to get this right in a way that it's not been done before. But I would argue probably probably twenty percent of our overall pain comes from the technology. You know, we can usually kind of iterate on the technology.
The hard things are, you know, when you've got unwell patients, how do you how do you make them? How do you reassure them this makes sense? You've got the healthcare professional that goes to see them day to day, how do you how do you reassure them?
You know, how do you make something that's usable for them to take into that patient's home and help them with and support and ensure that you know, they're comfortable with it? How do you then you know, make sure that the the more skilled nurses or the physicians over in the over in the kind of operations, you know, of that, you know, in the case of home health care are aware of what's going on and understand the context of what's being presented to them?
You know, how do you satisfy the financial director that any of this is worthwhile financially anyway? And it's, it's those is that narrative, you know, making those arguments across all those stakeholders, the system, let alone how do you actually implement something that you can get a kit from a logistics center to a patient's home, they use it successfully, and then it comes back and gets cleaned, you know, that's the workflow piece.
But, you know, bringing all that together is just a lot of complexity. And, you you try and take half an hour of someone's day in healthcare to go and to go and train them on, you know, how to use our solution perfectly. And you'll probably get about two minutes in a page will go off.
That's that we'll not see them again. And next thing you know, it's a support call where they don't know how to do something. And that's fine. We just have to make sure we accommodate that because these are busy people. And what matters is patient healthcare.
Dealing with technology is just not important, really. Yeah, I guess the technology needs to be kind of a seamless part of the solution that they doesn't even get in the way. So you guys are at this stage now, how many people did you say?
Seventy? Seventy ish people? About seventy now. Yeah. Plus a lot of subcontractors and focus on in the US as well, but about seventy full timers. Yeah. Okay. So you've you've you've grown very quickly over the past twelve, eighteen months, but you're still a startup.
You're still a small company in the bigger picture. And you're taking on huge incumbents in the health care industry, particularly in the US, I'm assuming. Yep. Have you got the sort of disruption advantage being the fast, nimble little guys? Or have you does the system and dealing with this impossible sort of Quixotic system, does that hamper you?
Or does that favor the bigger companies more that they've been around? They know how it works? How hard is this space for startups, I guess, is the question. I mean, I suppose going back five years when we first started this out, and we we told a few people that kinda knew a few few things about this space.
I mean, I suppose the the wording the statement that was made is we were too stupid to realize it was impossible, which I think was a pretty fair statement in retrospect is that, you know, there's nothing to see really easy about healthcare. I mean, we're obviously five years along now that we at least we've got something behind us that we can prove that we you know, we're trustworthy, and we can get this done.
The big company is quite like regulatory, and they, and regulatory has to exist for patient safety. But some of it starts getting a little bit you can you can see unless you're a big company is essentially impossible to comply with what's required. And that requires huge amounts of investment on teams of people that can can help manage that whole process.
And it is is, you know, whilst it is absolutely necessary for safety and an important part of building anything in this space, it does make it very, very difficult. However, you know, being a startup, you're absolutely right does give us a little bit of something that we can move quickly.
On some of the projects we've been working on over the last six months, in particular around some of the kind of events happening right now. We have had some of the world's biggest companies operating in this space have come to us to come and solve problems that we are, you know, we are well placed to solve because we're a small nimble startup, the kind of things that would take you know, three months of, of kind of team project management for us is six days of getting our heads together and throwing a few engineers at
it that are passionate about getting things done. And we brought together clinical expertise on top of kind of engineering expertise, with a direct line to manufacturing logistics and, and likewise, direct line to regulators that want to help us as well that we can kind of build relationships there.
And we can compress those timelines down. And that's been absolutely crucial in the last six months. We've become a bit of a go to, think, for those types of problems, which is kind of good that we can deliver on those. Yeah, it seems like you're, you're fighting a good fight, you're becoming maybe the flagship for particularly, particularly maybe for European startups trying to get into health, the healthcare market in the US.
And does it have to be the US, by the way, is that just the default that that is where the business of health care moves quickest? Or is it where the most returns are? Is it where what are the main reasons that the US rather than or is it just the fragmentation in Europe versus America?
This is something that really does come up a lot. I guess, the first of all, it's really the fragmentation allows innovation to thrive in small areas and then start growing and developing. You know, it's very difficult, you know, in a universal healthcare system that we have in the UK to get the early roots of innovation somewhere and then actually continue the momentum and keep on growing in a timeframe that's gonna start up compatible.
The US flips that on its head. And it means that we can, we can sell into one geography, geography, one region, because obviously, the economics are there to try and figure out how to either save money or make more revenue. And those two together are very potent for kind of figuring out, you know, how do we how do we sell solutions and co develop what those are in the US is very accommodating to those early, those early days of growth where things don't always go so
well, but you can figure out how to solve this. Yeah. Certainly, you know, in Europe, we want a very finished final perfect solution. That's quite difficult as a startup when you know, this space is so complex, it takes so long to develop this.
That said, it's very easy in the NHS to kind of get early, you know, early pilots. One of the one of the big downsides of certainly the UK is, is, you know, startups is a very famous concept in in healthcare of death by one thousand pilots, is that you have all these kind of great starts and actually show value on a lot of them.
But turning any of those into meaningful commercial kind of growth is actually pretty difficult. That said, we are getting huge interest from South America and Asia, Japan, Japan in particular, Indonesia around around those those kind of countries right now, I guess that aging, aging populations and health systems that are trying to figure out how to solve some of the problems around capacity.
The issue is up for us, obviously, it's more one of language that the US is English speaking. So it's quite an easy beachhead to go for that. And we are now in multiple languages. And so you know, it's easier for us to keep on growing that.
But, but the next step is certainly going to be I think Asia for us and further on beyond in Europe too. Interesting. So lots of geographies left for you. And just getting started, I suppose really, we question in from one of the delegates, Thomas Nourocchi, raised a question about the ethics issues in AI and medical tech, and specifically talking about bridging the gap between general research data and individual patient care.
I guess that's something that's something that's in a spreadsheet versus a person that's in a bed. Yep, absolutely. This is, this is something that I spent a lot of time trying to consider and reconcile where the ethics of this come into play. And I think it's fair to say, mean, stepping back from answering this question, one thing I've, I've kind of spent a lot of time trying to understand is that, you know, doctors, clinicians spend a lot of time working on ethics, the average kind of machine learning researcher doesn't do
anything really, certainly in no course I ever did, I ever step back and look at ethics, which is kind of interesting. I think that's something we need to address in our universities that people working in data science and machine learning, you know, certainly in healthcare really do need to understand what medical ethics are and why they're so important.
And the consequences here and regulators are trying to put in place some of the I suppose some of the foundations to try and you know, ensure that there's some degree of safeguarding there. But, you know, I spent a lot of time out in Silicon Valley working at scale and some of the big tech, you know, back in the moment to twenty twelve, kind of onwards, and seeing like how large scale social data was being used.
And some of it was quite interesting that, you know, there are a lot of techniques you can use to try and anonymize and identify the data and to a large extent, you know, that does, I suppose, solve some of the most acute problems.
But but that's, you know, when you're considering things like ad clicks and search engines, which are very different kettle of fish to the consequences of when things go wrong. I think this is as much a technical problem as it is a societal problem of understanding and, and really kind of truly recognizing both the value, but also the downside of this.
I think consolidating power into big tech is probably not going to be the best solution right now to really make this to make this you know, satisfactory to society at large. We obviously know Google has huge amounts of data on us already, this is a huge area of concern.
For us, personally, you know, privacy as much as about, you know, letting the patient know what we know, what data collection about them. And that fundamentally, you know, what are we doing with it? And that's a crucial part of this whole process, we can legislate and regulate and you know, all kinds of technical methods.
But at the end of the day, patients need to have awareness. And that for us is the most important part of what we do. And that's part of you know, that's wrapped into how we provide our platform to patients as they are given guidance on exactly what's being collected and why that's valuable.
And there's a very explicit contract there to allow them to make an informed decision. Yeah, it feels like there's a lot of things in tech right now that until the end users are actually the owners of their own data, it's going to be difficult for things to get better.
Stuart, I think that we're going to wrap it there. That's tons of interesting stuff. And we could be here all day chatting. There's a million different things we can do then, particularly if we open up the Star Trek Avenue. But we won't do that.
We're working on it. Any more than we have done. I do have a final question actually. Is there ever going to be a world where Current Health, I just buy a Current Health device off the shelf or I have a Current Health app on my phone.
Is that a likely future? We do see that as a future. We think that's going to be exactly where this all goes is that ultimately it becomes consumer tech, but it's part of day to day health care. So absolutely. And we certainly like to be the company that provides that to the world as well.
Cool, cool. Well, man, I hope you I hope you can make it be great to see one of Scotland's flagship companies taking on the world. It's great. We had Chris in last year on stage at Turing Fest and now having you in. So we're going be keeping a close eye on the future of Scotland's next unicorn.
And I appreciate that. Thank you. We'll you back next year to catch up and hear how it's all going. So thanks very much, Stuart. Great stuff. Thank you very much, Brian. Cheers. Okay, folks. So we're gonna we're gonna take a quick break now.
So we've had we've we've one more talk to come. We've got Mark McCloud, formerly the CFO at Shopify. It's good. You gotta watch out for that one. But also if you wanna grab a break now, go see go over to the booths and chat with some of our partners.
We've got Eisethel, FreeAgent and Digital and our old pals at administrate are waiting with open arms to hang out with everybody. I'll be back in fifteen minutes and our next session starts at three so we'll see you then.