Dr Andree Bates — Pharma AI, Saving Lives with Data & the HelloFresh of Medicine

Speaker:I'm Adam Walker, a biometrics consultant, and this is the Pharma Prescribed Podcast where leaders, innovators, and hidden voices in healthcare open up, no sound bites, no spin, just raw insight, one prescription at a time. In an industry driven by data protocols and pressure, we rarely pause to ask the human questions, what drives us, what breaks us, and what truths live behind the titles we wear.

Speaker:Dr. Andre Bates is a pioneering force in pharmaceutical ai. She's known for transforming how life science companies harness data to drive growth, innovation, and patient outcomes. As founder and CEO of Eularis, Dr. Bates has spent over two decades helping top tier pharma and biotech firms solve complex commercial challenges using AI with a PhD in neuroscience and cognitive science, and a career that began in medical writing and digital transformation.

Speaker:She's long been at the forefront of healthcare innovation. Dr. Bates lectures in MBA programs on health innovation and ai, and contributes thought leadership with work reflecting a rare blend of scientific rigor, commercial insight, and visionary thinking. Dr. Andre, welcome to Pharma Prescribed Today. How are you?

Speaker:Speaker 2: Thank you for having me on. I'm good.

Speaker:It's an absolute pleasure to welcome you today, and for our audience who may not be familiar with you, perhaps you could give us a little bit of background to yourself and how you got into this industry in the first place.

Speaker:Speaker 2: Oh, so as you said, I'm the founder of Eularis and Eularis Group.

Speaker:And I've been in pharmaceutical and life sciences, AI for 22 years, and a lover of future tech most of my life. I love all sorts of technology, and I got into this because during my PhD it was in neuroscience and in my lab we were experimenting with machine learning, so creating machine learning models to validate the neurological theories we had.

Speaker:At that time I didn't really think too much about, I didn't want to become an academic, I wanted to see an impact. And I think that's probably a defining part of my personality that I don't like doing things that don't have impact on real world impact on things.

Speaker:And so when I was doing my PhD. In neurology, it was basically the things we were looking at wouldn't have an impact on patients' lives for so long. And so that's why I didn't want to be an academic. I thought, I don't wanna spend 30 years to get to a real world impact. And so I looked for some way that I could combine medicine and technology and I didn't really know what that looked like.

Speaker:I saw a job advertised and I'm from Papua New Guinea originally, and I grew up in Japan. And so of course I speak Japanese. And so I saw this job advertised, which was asking for someone who was fluent in Japanese and had a PhD in neuroscience. And I didn't really know what it was, but I was like let's have a go.

Speaker:I applied and got it and that was in the pharmaceutical industry. And actually my very first area that I was working on was HIV. So it was really nice to be doing something that actually had an impact on real people and real people's lives. So I guess that's what got me into pharmaceuticals.

Speaker:Speaker 2: And then ai, it was really when I just over the years being in pharmaceuticals starting in digital actually I started in medical and then felt when I got headhunted to a job in Tokyo and I felt excited to go to Tokyo. 'cause I thought finally somewhere with good technology, but. At that time, it wasn't great technology in Japan like most places and I'd come from academia, so I had high expectations that companies didn't really meet in terms of the technology they were employing.

Speaker:So I thought, okay, this could, this job in pharma again, is gonna be great because they'd have great technology in Japan and they didn't. So I wanted to change that. And I am that sort of person that when I see something's broken, I want to fix it, which may be not a good thing at all times, and I really want to change.

Speaker:It wasn't my job to do this. I was actually doing a job in medical writing, but I really wanted it fixed. So just by chance, within a month of starting that job. I happened to be sitting next to the chairman of the company when a whole group of us from the company were catching the bullet train from Tokyo to Osaka.

Speaker:And he said, oh, you're the new person, how's it going? And I was young and too truthful, and I told him exactly what I thought was wrong with the way things were run and how I think it should change. , He looked at me and I thought, oh dear, I shouldn't have said that. And then he said, fix it.

Speaker:And I said, can I? And he said, yes. I'll give you whatever you need. Fix it.

Speaker:It was just the best opportunity. It was the most amazing experience to have the opportunity, to do that. And so that's was basically my first start at digitizing, which was a long time ago. I digitized the company then I got the reputation in the industry of digitizing.

Speaker:We did lots of world first with Glaxo actually digitizing things. I got headhunted again to another firm and did the same thing. Really got into digitizing and doing a lot of things that were really first in digital and healthcare. I really wanted to go in another direction with it.

Speaker:Because I like to push the boundaries and I was told that the company liked what I was doing, and we didn't want that. No one wanted me to change what I was doing. So I made the decision to leave the company and start my own company. And that was in Tokyo.

Speaker:That was in digital health. After that , my first client was Pfizer in New York who didn't seem to realize I was in Japan. I was doing a lot of content marketing. I moved to London because I had clients in Europe as well as New York to begin with, as well as Japan.

Speaker:I was constantly going to New York all the time. I moved to New York for a short time.

Speaker:Speaker 2: Anyway, I was doing that for a while then I saw other issues, and this is probably a glitch of my personality, which may not be a good thing where I'm always liking to solve new challenges. And one challenge I kept seeing is that in digital we could measure everything.

Speaker:So I could say exactly how many prescriptions were increasing due to a digital program, for example. But I saw in the rest of pharma a lot of it, and a lot of people used to say, in those days, 50% of my marketing's wasted. I just dunno which 50%. I thought, why not? Why can't we fix that? So I then started Eularis to really apply mathematics to pharmaceutical commercial decisions.

Speaker:And very quickly I realized that I needed a non-linear math, and then I remembered my PhD and had some value. I remembered machine learning. So then I worked on building, which we did a platform that was basically predicting pharmaceutical market share based on collecting a lot of data from doctors and all the data, all brands in a category.

Speaker:We had often a hundred percent accuracy, but over thousands of brands. We had 94.8% accuracy. So our clients used to call it 94.8. So we were doing that and that was very successful.

Speaker:Speaker 2: But one client who had an ultra rare disease came to us for forecasting and they said, can you do some forecasting?

Speaker:And we said, sure we can. But wouldn't it be better to use artificial intelligence to find those patients rather than just do some forecasting? And they said, how could you do that? And so we came up with a way, and we found all these undiagnosed patients that they were, the children would die if they weren't treated by the age of 10.

Speaker:They would die. And so we found these, a lot of really young toddlers. I was speaking at a rare disease conference in Boston earlier this year, the World Orphan Drug Conference. It occurred to me that it was now 20 years this year since we did that project. That means that those children that were really young toddlers would be in their twenties now.

Speaker:And alive and living, normal, productive lives. Because we did that project and that changed everything for me. It just made me feel like now that's impact. If you're doing something that's saving real people's lives, that's what I want to do. Probably not for the best profitable reasons, because, having a platform is a much more profitable, steady thing to do.

Speaker:Solving unsolved problems is a much more costly, less profitable thing to do. But it just made me feel so, like we had a mission and a purpose that was changing the world. And that's what we've continued to do.

Speaker:Wow.

Speaker:There's so much I would like to unpick there and really it is such a remarkable story of.

Speaker:Curiosity and challenge, and you come back to that original point, which was you wanted to make a difference, but you wanted to see that difference in your lifetime. Yeah. And I really do resonate with that principle, as much as appreciate what that must have been for you, including all of those changes in culture and location and parts of the world, the sorts of things that you had to personally presumably have overcome in order to be successful in every one of those different geographical locations of the world.

Speaker:Speaker 2: I think when you grow up in several very diverse countries, you become very adaptable. I'm quite easygoing about where I live. Any country I can adapt to really, obviously I don't wanna be in the middle of a war zone, but, it's living with people from a different culture is such a great thing to do and you just realize we're all the same.

Speaker:Are we or are we just influenced by different opportunities and dare I say, without going down that rabbit hole geopolitical challenges that we tend to be faced with. It sounds like you are flexible not only on a personal level, but on a professional level and given all of your experiences and languages that you have to your disposal, that you are able to be a chameleon in each one of those different situations.

Speaker:Is that fair? Is that reasonable?

Speaker:Speaker 2: Yeah, I guess so. I like to immerse myself in whatever culture I'm living in yeah.

Speaker:And that's wonderful. Thank you. Thank you for explaining that and leading into the conversation of your why. I think that's really what we've established there.

Speaker:So now moving on to some of the implementations of technology that you've been privy to and that you've been directly involved in. I think we have a similar sort of experience in the fact that when you entered this industry, it wasn't digitized. You were responsible for doing that digitization.

Speaker:I worked in many systems and platforms that also did those things for clinical research, drug trials, and medical information. So are we really at an inflection point today? Have things changed to never go back again?

Speaker:Speaker 2: I don't know, but I think there's been many inflection points in the last 20 odd years in AI in general.

Speaker:We had. I think the really big ones were 2014 when gen AI was invented. And it was invented by Ian Goodfellow, who was a British data scientist. And I think that was huge. Then we were able to start using, generative AI in lots of different ways, including synthetic data, with rare disease.

Speaker:So that was a really big change. And he started with GaN, which is generative adversarial networks, but there's so many different types of gen AI now, which are all the type of machine learning, of course. I think the next inflection point was 2017, where we had more data generated in 2017 than the entire history of mankind up to that year.

Speaker:The data started expanding exponentially then every year after that. And so that's. Great for AI or not great, depending on the way you view it. But it gives you a lot more data. In the early days , we had to go and collect data and it was a big expensive process. I'm not saying it's not a big expensive process today, but there's still a lot more data now than we ever had in the early days.

Speaker:So it means that we've got a lot more clarity in what we're doing. And then of course the , the transformer models came out, I think it was 2018, and then we had the GPT models from 2019. And now of course, LLMs are huge, although I don't think LMS will last that long. I think in the next five years they will be superseded for sure.

Speaker:Not for sure, who knows, but I feel like , they're great. They do a lot, but I think there's a lot more technologies that will come. After that. But yeah, it's been interesting.

Speaker:I'd love to hear more about your perspective on, on LLMs because for the lay person, and I'm including myself in that as much as many of our audience we're just getting to grips and getting used to having this in the palm of our hand in mobile phone technology and tablets.

Speaker:And bringing back to that original point about getting hold of patient data in the early days, that was challenging. Today, is it still challenging, is it still difficult to get through those data privacy laws that we have , in various different locations? Is the data really as easy to come by

Speaker:as people are suggesting, because I also have worked in real world evidence and rare disease. The impact of that for people with rare diseases and people living with rare diseases and children and relatives is enormous, isn't it?

Speaker:Speaker 2: Absolutely. The data we work with that is patient data is anonymized, so you don't know all of the personally identifiable data is taken out.

Speaker:However, you cannot take any patient data out of a country, so you have to spin up a cloud instance in that country, and then you have to use a federated learning approach to be able to basically combine that data with data from other countries because it's not allowed out of that country. So that's interesting.

Speaker:I think it is possible to get from hospitals now and, we've done deals with hospitals. A lot of people have done deals with hospitals and hospital networks to get data. But again, , you dunno who that patient is. You've got a lot of information about them, but nothing identifiable.

Speaker:Speaker 2: We've done a lot of rare disease projects where like one we did for a third line cancer product that was so rare to find people who were still alive by third line. So the company had both a lot of trouble in their clinical trials because they couldn't find any third line patients for that.

Speaker:But then also when it eventually launched for their marketing, which doctors should we go to? 'cause all doctors, most doctors, we see the patients are all dead at second line. For that one, we could get the patient data. What we did actually in that one was interesting. We looked at EHR data electronic healthcare record data of patients that had that kind of cancer and died at second line and ones that had that kind of cancer and survived to third line even if they were already dead.

Speaker:Now, of course, we dunno who they were, but we've got information on them then we ran algorithms to compare the two groups so then we could create a high probability of someone who's alive at second line continuing on to be alive at third line. What we did from that is, we'd love to know who they were so we could get to them, but the only way we could do that was to get claims data, which didn't have, the patient into either, but it had the doctor ID so we could triangulate.

Speaker:The claims data with the EHR data still dunno who the patient is, but we know who the doctor is. So then we could go to that doctor and actually really pinpoint, these are the doctors that need to know that we have a third line cancer drug for this condition because there's a high probability that, one of their patients is gonna make it through to third line.

Speaker:Yeah.

Speaker:But for those of our audience that aren't familiar with first, second, and third line, would you mind just explaining that I'm not familiar with that myself, but more broadly, just from a very simple standpoint, what does that mean?

Speaker:Speaker 2: First line cancer is the.

Speaker:First time it comes out in a person and then, oh, sorry, second first line treatment. So when you have cancer, you get treated and if you have FirstLine treatment, it's basically your cancers, fairly early stage. The worse it progresses, you can then go to a second line treatment and then if that doesn't treat it and the cancer still progresses, that's when you get to third line.

Speaker:There aren't that many I dunno now, I haven't looked at it recently, but a lot of cancers, people die at second line. In this particular cancer that was the case. There hadn't been a third line treatment and that's why they put it out to save lives of people who actually survived to third line.

Speaker:So it is, basically the severity of the cancer.

Speaker:Thank you for explaining that. I think that's very helpful. The triangulation method , you described with bringing together the claims data with patient data all being anonymized.

Speaker:On the one hand, possibly for many of our listeners that might be reassuring to know they can't be identified. But at the same time, there's some very complex programming isn't there, around trying to match this data together in a real world data set. And that is the thing that I've also seen firsthand that is moving forward at some pace,

Speaker:Speaker 2: and if you think about the size of that data, the claims for the US for example, or the EHR data for the us 'cause companies like Epic have something like 80% of that data. So it's huge. You've got, hundreds of millions of people's data. To be able to do that, the only way you can do that is with AI algorithms, machine learning kind of algorithms, because it's just so massive.

Speaker:You couldn't actually sit there, a person with an Excel spreadsheet and try and figure it out.

Speaker:So is this a skill that you are now embedding within Eularis? Are you using highly qualified teams of people to be able to perform these data analyses?

Speaker:Speaker 2: From the beginning I have done that, yes.

Speaker:I've got people from really good AI companies and when I first started, I was really looking for the top data scientists in the world. And data scientists are , very curious people like me, but. Better than me at this. They do a lot of things like they play chess and they do competitions of challenges, of mathematical challenges.

Speaker:There's a data science competition that's been around for a long time. Google bought it actually. But the top guys on that, I approached the top 20 and I actually got 13 of them. And I did get people from Google and I've still got them actually. And I did ask them a few about 10 years ago, I said, why are you with me when you could be at this huge company?

Speaker:And they said, it's because when you're working at a huge company, you've got one problem that you're working on for years and years. Whereas with us, it's. Different interesting problems all the time. And that actually is, something that data scientists, as a personality type in general they like to have really difficult, at least the good ones, really difficult, challenging problems to solve

Speaker:and plenty of them presumably.

Speaker:, By having identified that sweet spot of brilliant people, are there other characteristics that they have that perhaps the rest of us don't? What is it that differentiates them from the rest of us other than just being laser sharp focused on data,

Speaker:Speaker 2: So there's not just data scientists that you have in a team.

Speaker:Of course. So data scientists tend to be like that. They tend to be, a super brilliant but also, so they usually have a PhD in maths physics or computing. And then they have to have specialized in ai, but. Then you also have your big data engineers. You've got your DevOps guys. So there's, your full stack developers, there's a whole tech team involved in any AI project.

Speaker:And of course the scrum master who keeps it all going. And I tell you, when we started with Scrum, I don't remember when it was probably 15 years, maybe longer ago. It changed my life. Agile is such a great way to go. Before that, we'd done things in a kind of waterfall approach, which, was very slow.

Speaker:And with AI you really need to be agile. So that's amazing. I've got someone who was the head of AI at Glaxo. So I've got people who are very deep in the pharma and AI space as well.

Speaker:Speaker 2: So we do, 50% of our work is doing AI strategic blueprints.

Speaker:50% of our work is doing AI strategy for pharma and life science companies.

Speaker:And we've been doing that since about 2011. And the way that came about before that, all we would do would be solve problems. That pharma or life science companies would come to us with ai, we'd solve it with ai. But at that point, this global CTO of a top pharma company came to me and said, we want ai.

Speaker:And I said, okay, fine. What do you want to do with it? And he said, anything you like? And I said, okay, what business unit are you thinking? And he said, any business unit. So I said, okay, we need a strategy. Before that, I actually said to him, why did you come to us with this? And he said, the global CEO realized that we need ai.

Speaker:He's given it to me and I'm giving it to you. So that's. Where we got the idea to do strategy because we realized that there's probably a lot of people in that situation where the CEO is a visionary and sees that AI is important, but they don't understand it. They don't really know where it fits. And they mistakenly assume that the CTO traditionally would know AI and it's actually a very different field from it, but they give it to the CTO.

Speaker:So what we really need to do is see how a company can leverage artificial intelligence to solve business problems, get them to where they need to be. What we start with is looking at what is the company, CEO, and board trying to achieve in the next year, the next three years.

Speaker:Then we look at the blockers to get to achieve that. So IT business unit by business unit, we look at what they're doing, what their pain points are. How that business unit contributes to the overall objectives of the company and where it's being slowed down or hindered because of, some clunky process or something not quite right.

Speaker:Then we think of all the ways that AI could actually be leveraged to solve that. We come up with hundreds of different ways, but then we have to think, okay, which of these, what is the few, what is the top 10 or the top five that are going to give the most impact?

Speaker:We didn't do this in the very beginning, but now we've got a team of financial modelers. So I've got guys who've got masters and PhDs in financial modeling with private equity backgrounds. What they do is they model the company and we model what would happen to the company with the ai.

Speaker:With 22 years experience, we've got a pretty good idea of. What impact different kinds of AI will give. Now, when we did this all those years ago, 2011 was the first one. There weren't really AI vendors out there. It was really few and far between. So because we'd been building ai, the top solutions that we came up with that they want to go ahead with, we built for them.

Speaker:Nowadays there's a lot of amazing vendors and I showcase them on my podcast. That's what I like to do. And so when we do a strategy, probably about 50% of the ideas that we come up with of how to solve problems, there's an off the shelf solution. So the other 50% there isn't because no one else has had the idea apart from us.

Speaker:So for those 50% that there are solutions out there, we do an in-depth vendor analysis. We look at the claims of each of the vendors in that space. We look at their data privacy, their security. We look at what their clients say, and we really do an in-depth analysis of them. And we make recommendations of, this is the problem for a quick win.

Speaker:There's a platform that exists. We've looked into them. They're not just a pretty face. They've got good ai, they've got good results. Not just what they say, we've tested it and this is who we recommend.

Speaker:Speaker 2: That's great because that can get you some really quick wins and cost effectively, because most of those companies are venture capital funded.

Speaker:They are, basically selling it to a lot of people. And you can get up and running really quickly and cost effectively. So that eases things that are, time consuming and process driven a lot of the time. But if it's something that you want to add unique value to, you want something that's quite different that no one else has, that is going to give you a competitive advantage.

Speaker:We haven't found vendors that do that. , In that case, we build it. The other 50% of the business is essentially building ai. They're the main two areas.

Speaker:Speaker 2: We also, of course do some training because I feel like when we do a strategy, there's a bit of change management involved when people have to, learn a new process.

Speaker:Although when we're doing the strategy, we are looking at people's workflows showing them that it's not that you're learning a whole new thing. Look what this is saving you in your day-to-day work.

Speaker:I did one for a life sciences agency this year. As we did a strategy as well as I was doing a training. And when I was doing the training, one of the first things they said to me before I even started the training is. This isn't gonna work because LLMs, we've tried them all, they're rubbish.

Speaker:And I said, okay, let's look at that. So we looked at their actual use cases and said, what is a real use case of a, something that you're doing, that you're spending, a few days a week on that's taking up a lot of time and let's look at what you did. They went to one of the LMS and there's some really good specialized ones, by the way, for life sciences, and they put in these very basic prompts.

Speaker:So they didn't actually, describe it particularly well. We did a prompting exercise and then they did again, they were quite impressed with the difference. And then I'd already done pre-written prompts for them because I'd already asked them for their use cases. They were amazed at what they actually could get out of it.

Speaker:I still think for LLMs at least, you need a person in the loop to check everything because you never know. But we have been able to achieve really strong accuracy with those when we put guardrails in them. But I think that whole showing people what time saving AI can give them in their role that they do now is where that change in people's mindset comes about.

Speaker:Speaker 2: I think there's a lot of fear also around AI and people are worried that they're going to lose their jobs. And we always advocate, we can save God hundreds of thousands of hours a year, but at no point do we say. Get rid of anyone because actually humans add a lot of value.

Speaker:And so if we can free up the humans to do much higher value tasks, then the role is pivoting for sure. You need to be able to know how to work with ai, we have such specialized people in all departments, in all business units, that we really still need those people

Speaker:i've never had juniors in because I've already gone for a really high level, people who can hit the ground running. But with the big consulting firms that, hire people outta university and train them. They're not hiring the juniors anymore because they can use AI to do that junior role.

Speaker:I think in pharma, we need to have our specialists that come up through the ranks and learn, medical affairs, regulatory, all of those things. AI can speed up those roles and do amazing things. We've done a lot of projects in medical affairs and regulatory, but you still will always need those specialists.

Speaker:For sure.

Speaker:I'm glad you've said that, and thank you for elaborating on that , Dr. Andre. Because it's a point that I've made many times with alumni at my university and medi many young people coming into this industry. I would hate for them to think that there is no longer an entrance into this industry.

Speaker:One of the reasons why I originally started this podcast was to enable people to understand how we can push open the doors. To the next generation because you and I are not always gonna be sat in these seats doing the jobs that we're doing with the things that we've learned along the way. Your experience is remarkable and I would love to hear some reassurance beyond that for that next gen that are coming through, that are listening and following the podcast.

Speaker:Because it's more than just reinforcing that there will be jobs and there are humans in the loop and there are higher level roles. As you said, you've got to the position where you have, because you've learned through doing and built that experience up over a period of time. Yeah. What message can you put to that next generation of life scientists coming into this industry?

Speaker:Speaker 2: I think going to pharma more than consulting, but I think with pharma, yes, it's always gonna be needed and I think that they need to take those positions and work through, because that's. Always gonna be a critical skill in pharma, I believe. But I think in the big consulting firms, it's going to change for sure, and it's going to be harder and harder to get the high level consultants in the future if people aren't coming through the system.

Speaker:And dare I say it, you indicated that you bring in oven ready, high level talent. Yeah, very high expertise. Is there a flip side to that? Do you ever consider bringing in younger people, fresh to the industry? Is there a reason why you've gone down that particular route? I'm sensing that there might be,

Speaker:Speaker 2: I think because I went down that route because I was small, so I needed people who could hit the ground running.

Speaker:We do mentor and we do a lot of mentoring work with with students. And, I'm lecturing at Imperial in January. We do lots of lectures as well, but we do mentor and. Funnily enough I've got a guy who is doing his PhD that I'm going to bring in as an intern next year, and I met him through an Uber driver from Uganda.

Speaker:I was talking to the Uber driver and he said his cousin from Uganda was doing his PhD in data science at Imperial, and he's got a mentor from DeepMind. And he said he's, looking for internships. I was like send me his resume. And he's such an intelligent kid, so we are bringing them in, but I wouldn't leave them to just run on a project.

Speaker:They're more coming in as interns to learn and shadow.

Speaker:Yes. And I think that makes sense and I'm delighted that you made that point. Serendipity brings people together all the time, doesn't it? Yeah. And you'll find it as much as I do. . Sometimes you just don't know where those.

Speaker:Geniuses or that magic is gonna be unearthed.

Speaker:We've touched on many aspects of your experience and it's fascinating to hear from someone who is really at the absolute sweet spot of AI and clinical research today. Your podcast, you've done several hundred episodes now. Is there one piece of advice or learning that you've taken from that you could share with my audience?

Speaker:Encapsulated in a very brief headline? If you could

Speaker:Speaker 2: brief headline. That doesn't sound like me.

Speaker:Do you want me to give you my vision of the world? Please. I would love

Speaker:to hear your vision of the world, Dr. Andree,

Speaker:Speaker 2: from lots of podcasts, it sounds like sci-fi. And just as an example, I was in a meeting in Virginia this year and I went to the airport and there's a whole heap of issues around Newark Airport and not enough traffic to air traffic control.

Speaker:All the flights were being, sent to LaGuardia, and LaGuardia was shutting down. I thought I'd be in Virginia for an hour's meeting and I'd be back in New York, no problem. But I ended up being in the airport for 12 hours with my fellow passengers on the flight. I was entertaining them about AI and pharma and AI and healthcare for many hours.

Speaker:And they were all saying, it sounds like science fiction. I was like, but this is what we have now. This is not science fiction. So let me give you some examples. And these are some of the podcasts that have stood out, even though they're not all ones I've done in the few last few weeks, but.

Speaker:Speaker 2: There was one I did maybe two, two or three years ago with the Chief Medical Officer from Grail, which is a diagnostic company.

Speaker:They use AI to diagnose cancer. They take a blood draw like a tube of blood and they can predict and they did clinical trials on 330,000 people. They can predict over 50 types of cancers in people with no symptoms. I did that podcast and I found that amazing and it did make me think we are going to be more focused on AI for diagnostics in the future.

Speaker:I don't know if doctors will be doing it as much as. AI because the doctors only have their medical school and their lifetime of experience, whereas AI has millions of doctors' lifetime of experience. But even more, it resonates with what we talked about the very beginning, about Impact. A few months after that podcast came out, I got an email from someone and they said you saved my life.

Speaker:And I said, how, what are you talking about? And they said, I listened to that podcast and I listened to it a few times and then went and got the test found I had cancer and then it's been treated and I'm now cancer free. That was amazing. I think that what Grail are doing are really interesting, but there's also a lot of other amazing companies.

Speaker:So there's a guy, Francois Gand, who. He had a lot of tragedy in his family. I've had him on my podcast twice and I'm actually gonna get him on again because he keeps pushing the boundaries.

Speaker:Speaker 2: Have you heard of Elon Musk's Neuralink, that implantable chip?

Speaker:I haven't. Okay. So Elon Musk for quite a few years now has been implanting computer chips in people. It hasn't gone well. It's gone. Okay. But they've had a lot of problems with the wetness of the human body and computer chips, degrade with the wetness. So I don't really know where it is now.

Speaker:I haven't looked recently, but Francois Gand basically has done that in a way that is non-invasive and works. So what he wanted to do was to allow people. Who couldn't speak, who were totally paralyzed, who couldn't even move their eyes to communicate because someone in his family was like that. He is a computer scientist, data scientist.

Speaker:He developed something that is a plastic headband that just sits on your head and within an hour people can be communicating. Even if they can't, they're completely paralyzed. And the first patient that it actually worked in, he said the whole hospital room were in tears. The doctor was in tears, everyone was in tears.

Speaker:But now it takes an hour of setting it up and calibrating it with someone and it's working. I think it's said ine quite a lot of the American hospitals now the network hospitals are actually. Buying it and using it and it's quite a game changer. I had him on again and he has now developed it further so you can understand the emotions that person is feeling as well, which is incredible.

Speaker:And he's done something else now. So I dunno what the latest thing is. I haven't done the third podcast yet, but that combined with another guy I met who is the founder of a company called Cortical Labs in Melbourne, in Australia. And what he did is interesting, but has lots of implications. So he has taken human brain cells and grown them on computer chips and he then taught them to do things and they were superpowered because the brain.

Speaker:Was empowered with the computer chip, , and this is a single brain cell, so this single little brain cell and, but he actually keeps it in a box, like a briefcase size box that pumps blood to it, pumps oxygen to, it gives nutrients to the brain cell because of course human brain cells are living and they need blood and oxygen and nutrients.

Speaker:So he gives all of that. And so he is got this sort of artificial body that is feeding the brain cell. But what he discovered is the electricity that the brain cell gives off is astronomical. He then plugged this into computers and he found he could power computers on a human brain cell. So the CTO of Amazon found out about that and he flew to Australia and met with him and.

Speaker:Amazon we're thinking of him powering their server farms with human brain cells. Now, if we combine that thought, which is fascinating and also very ecological, if we only use 10% of our brain, imagine if we could harness the power, if one brain cell can power a computer we don't need all this electricity from outside sources.

Speaker:We're self-generating electricity.

Speaker:Speaker 2: So I was thinking, and I'll talk to Francois Gand about this, but I was thinking, I wonder if Francois or someone like him could actually create an external device that we just have on our head. Francois has this plastic band that actually conducts the electricity.

Speaker:From our brain into things. Maybe one day we could actually power our own digital twin from our own brain, which would change medicine and change everything, change pharmaceuticals because pharmaceutical companies would then basically identify the molecule and provide the active ingredients.

Speaker:You would test the right combination of those on your digital twin. So it's not impacting your body, its impacting your digital twins body, which is identical to your body, but a digital version which we have now already, but they have to be run on frontier and huge quantum computers because of the power required.

Speaker:Even if we couldn't power ourselves, if we've got one, when we get to the point where we've got our own digital twin, we will have truly personalized medicine and we will be able to, adapt a drug for our physiology. And as , most early stage trials are on men, not on women.

Speaker:So it's now we've got about 40% women in the phase three trials, but not the phase two trials. But it would be great if everybody had a personalized formula of that drug for them that actually worked. And of course, we're already doing many projects where we can look at sub cohorts of patients and know which patients, which particular profiles of patients will respond better to a drug than others that have got the same condition.

Speaker:But just imagine. Pharma could, if that case comes about, become a bit like, have you heard of the brand HelloFresh?

Speaker:I have, I'm familiar with them. They deliver boxes to your door and food, don't they? No. And all you have to do is follow the menu.

Speaker:Speaker 2: Yeah. So pharma could almost become like a HelloFresh of medicine where they deliver the formula for the molecule and they deliver the ingredients and you put it into your 3D printer, which, we've had 3D printed drugs since 2015 that have been FDA approved.

Speaker:Spiritan was the first. So you throw it into your 3D printer. First you get it, test the molecule on your digital twin, then you've got the right formula. You put that into your 3D printer or whatever. There's actually another technology of a guy I had on the podcast that actually is like 3D printing, but it's more scalable and cheaper.

Speaker:It's gonna change the face of pharma completely. And medicine, we're already now with companies like Grail being able to predict cancer when we can't even see any symptoms yet. It's a very interesting time.

Speaker:Your insight is beyond remarkable.

Speaker:Dr. Andre and I cannot thank you enough for sharing some of those thoughts and musings. You've got me thinking, and I'm gonna go away from this conversation thinking a lot more about what might be possible and indeed what my perspective on the next three to five years might look like. Possibly as many of our audience will at this point in the conversation.

Speaker:I'd like to finish with a quick fire round. So what is the one piece of advice you would give to your younger self, Dr. Andre?

Speaker:Speaker 2: One piece of advice. Let me think. I think I would've given myself a lot of advice because I did so many things. I guess maybe. Trust the vision, but be patient because when I started in 2003, I could see where AI would end up, but maybe not where it would end up.

Speaker:I still dunno where it's going to end up. But, back in 2005, we're finding undiagnosed patients and doing things like that. I guess what frustrated me is how slow the industry moves. If you are doing a lot of this work, you need to be a little bit patient. At least I had to be.

Speaker:Because lives are at stake and we've gotta be very careful. But I think, trust the vision, be patient and be kinder to yourself during setbacks. It's a marathon, not a sprint.

Speaker:I love that. What are the top three qualities you value most when building one of those teams?

Speaker:Speaker 2: One of them, which is the intellectual curiosity. Actually, although I have gone for domain experts, my top thing is intellectual curiosity over domain expertise. Expertise matters. I have experts, but they're all people who are genuinely curious and ask why and what if.

Speaker:And I believe it. In many ways, it's far more valuable to ask those sort of things. If you think through to what AI can do now, still it's learning from what's been done. It's not asking what if. So the answers you get are all from things that have happened, and that's something humans bring to the table that AI can't, which is that vision and that.

Speaker:Curiosity of what if. So I think that's another one.

Speaker:My second one would be I dunno if I should say this. Okay. Patient centricity. That's genuine. I have worked with companies that on their website that says patient centric and then you work with them. You say, what patient centric things are you doing?

Speaker:There isn't really very much. So I'm looking for people who really care about the weight of responsibility that we have with what we do. If our AI is identifying a patient with an ultra rare disease that's been suffering, and brings life-saving treatments. These aren't just data points or revenue opportunities, they're human lives.

Speaker:So the team members that make the biggest impact are really those that are thinking. What serves the patient best. And even when it's commercially complicated or technically harder, so it either drives you or it doesn't.

Speaker:Then I think the last thing would be comfort with ambiguity and pioneering.

Speaker:I can tell you a story that is a little bit funny. When I first started, I was at a dinner party and someone said, what do you do? And I said, I work in ai. And they said, oh, that's very controversial. And I said, really? Why? And they said artificial insemination as a concept in self. And I was like.

Speaker:Oh no, not that ai, artificial intelligence. So people didn't even know what it meant back then. So I think that you've gotta be comfortable with ambiguity and pioneering. And particularly with AI and pharma, we are often in uncharted territory and there's no playbook for, how to explain black box AI to a regulatory agency or, things like that.

Speaker:So I like people who can not just tolerate ambiguity, but thrive in it and, start working on a problem before we really understand all the requirements. I'm a pilot, I apply airplanes and helicopters, so I think, someone who can build the plane while they're flying.

Speaker:It we've never done this before, but , let's do it. I like that kind of thing.

Speaker:That's fantastic, and thank you for elaborating on that particular point.

Speaker:I'm curious to know, Dr. Andre, you may have just touched on it, but what is your favorite thing outside of work?

Speaker:Speaker 2: It's interesting. I'm so passionate about my work that , the lines are quite blurred as my friends will tell you, I went away with a friend of mine for a weekend, maybe a year and a half ago, and the whole weekend she and I were playing with ai.

Speaker:So I don't know. We, I'm very fascinated by that. But I also do love sports where you have to think and that you don't have time to think about other things. So I love flying. I started flying airplanes when I was about 17 or 18, I only got my pilot's license then, and I was flying all my life because my father was a private pilot.

Speaker:He was an engineer, but he had a pilot's license. So from the time I was really young, I was sitting there with my father. I got my helicopter license in 2011 and I love both of those. I went sailing in the Greek islands last year with a friend of mine and it was just us and we were sailing and the weather was amazing.

Speaker:But it was great because there was no wifi. It's amazing when, you can't check your email. And, you dunno what's going on with work, how you can actually turn off. You have to focus on your navigation. You have to focus on the wind and what you're doing. So that was great. There was another holiday where I was canoeing up the mekon delta in Vietnam and staying at villages.

Speaker:Whatever it is, doing something that takes you out of your, constantly thinking about work, but also there's something about being in an unfamiliar environment that sharpens your observation skills and reminds you that there's, infinite ways to approach any challenge.

Speaker:I often, if I go sailing or if I fly to another country or something, I come back with fresh perspectives and somehow, I don't know, they seem to find their way into how I think about pharmaceutical innovation. If you're building AI or navigating a foreign location, it's understanding patterns and respecting context and remaining curious.

Speaker:You sound so fascinatingly curious about every aspect of your engagement with the world, with life, with everything.

Speaker:So finally, what is your number one golden rule in life and in business, Dr. Andre?

Speaker:Speaker 2: I think the top rule would be to do work that matters with people who matter in ways that matter.

Speaker:Wonderful. We've covered a broad range of topics today. You have been very generous with your time, Dr. Andre, and the storytelling and indeed sharing your why. I know this episode will resonate so loudly with our audience, and I cannot thank you enough for taking the time to just tell your story today and share that so honestly and generously with me.

Speaker:For any of our audience, if they want to make contact with you, what's the best way

Speaker:to make that contact.

Speaker:Speaker 2: Um, LinkedIn, if you just put in Dr. Andre with two Es Bates you'll find me and I post. In fact, today, every Wednesday I put My newsletter out where, I discuss new topics in AI and life sciences.

Speaker:So that's the best way you can message me there.

Speaker:Thank you so much. As I say, it's been an absolute delight and a pleasure to welcome you onto Pharma Prescribed Today. I can't thank you enough for taking the time and I hope to continue the conversation at some future point. That would be amazing.

Speaker:Speaker 2: I'd love that.