Latest / Hardiness with Dr Paul Taylor / How Data-Driven Healthcare Could Revolutionise Mental Health and Neurodegeneration with Nawal Roy

Transcript

How Data-Driven Healthcare Could Revolutionise Mental Health and Neurodegeneration with Nawal Roy · Hardiness with Dr Paul Taylor

0:10
Speaker 1: Navel Roy. Welcome to the podcast Pleasure to be Air.
0:14
Speaker 1: Paul so Maya.
0:16
Speaker 2: You're basically an economist and a trader, right, and you've
0:21
Speaker 2: worked with McKinsey, Credit, Swiss, all of these guys in
0:25
Speaker 2: sort of economics consulting.
0:27
Speaker 3: What the hell are you doing in healthcare? I asked
0:32
Speaker 3: that question again and again. But if I could say,
0:38
Speaker 3: a simple thing that I learned from economics and finance
0:41
Speaker 3: was a commitment to data and how data is a
0:45
Speaker 3: massive enhancer in terms of understanding that field. And when
0:49
Speaker 3: I started HALLMSK, I realized that how healthcare overall was
0:54
Speaker 3: missing from data perspective to many expertise and too many opinions,
0:59
Speaker 3: but not enough data. So my whole take was, if
1:03
Speaker 3: I could detail for one second, is in finance.
1:07
Speaker 1: There is a product called Bloomberg. Bloomberg is the ultimate
1:10
Speaker 1: data infrastructure on which most of the finances of decisions
1:13
Speaker 1: are made. Yes, having that architecture of the high quality
1:18
Speaker 1: data was my goal post to build that for healthcare.
1:22
Speaker 2: Interesting, and you picked, but I think the two areas
1:28
Speaker 2: of healthcare that are the most underperforming. Right, If you're
1:33
Speaker 2: going to go in and fix something, and you went
1:35
Speaker 2: straight to you went straight to what I think are
1:37
Speaker 2: the well not the most broken. So you let me
1:41
Speaker 2: give a little bit of context on that. You picked
1:44
Speaker 2: mental health and neurodegeneration. So basically everything to do with
1:48
Speaker 2: neuroscience and mental health I have often said is it
1:54
Speaker 2: performs at a shockingly poor level in terms of our
1:58
Speaker 2: understanding and our into vengeance. And there has been also
2:03
Speaker 2: so much I can't think of an area of health
2:06
Speaker 2: care where there has been so much money invested in
2:10
Speaker 2: potential treatments that have ultimately feeled miserably as your degenerative diseases,
2:17
Speaker 2: right and emily combine the two of them. The amount
2:22
Speaker 2: of wasted money, in the amount of paucity of evidence
2:26
Speaker 2: of the treatments in these two areas, it's got to
2:29
Speaker 2: be more than everything else combined in health care, I
2:32
Speaker 2: would imagine.
2:33
Speaker 1: I mean, there's no objective her to say that, but certainly,
2:36
Speaker 1: you know, if you go by the history of overall medicine,
2:39
Speaker 1: certainly there has been a multiple failure in every other areas.
2:42
Speaker 1: But having said that, I would say neuroscience is the
2:45
Speaker 1: last frontier in a to understand overall health care, or
2:49
Speaker 1: to certainly understand the medicine and what we know of
2:54
Speaker 1: neuroscience is the last super complex, super large problem that
2:59
Speaker 1: is still not touched or has been only a scratched
3:03
Speaker 1: the surface. But in ten years from now it will
3:06
Speaker 1: be as good as what was the state of oncology
3:09
Speaker 1: thirty years back. So I see it from the perspective
3:13
Speaker 1: of what is the value add that I can build
3:16
Speaker 1: it so that the future of nuto science can be
3:19
Speaker 1: truly achieve the full glory. And my analysis is exactly
3:25
Speaker 1: where you are going after is when I started looking
3:28
Speaker 1: at the overall healthcare sector, I really found that the
3:30
Speaker 1: mental health is the largest uncapped, untouched you know, sexy
3:36
Speaker 1: beast that is not has been tamed, Okay, yeah, And
3:41
Speaker 1: every component of that is hard. Evidence is hard, science
3:44
Speaker 1: is hard. Payment problem is hard, stigmaized. Their access problem
3:47
Speaker 1: is there every component of there. And the underlying architecture
3:51
Speaker 1: of solving it is how do I solve the evidence problem?
3:55
Speaker 1: And the evidence problem is a massive data problem, and
3:58
Speaker 1: that data architecture is where I decided, Billie. So you know,
4:02
Speaker 1: it has been ten years building and director, but now
4:05
Speaker 1: we have truly built on which the future of neuroscience
4:07
Speaker 1: can be really have a crack at solving it.
4:11
Speaker 2: I want to go into the data and and and
4:13
Speaker 2: how you number one, how you gather the data, and
4:17
Speaker 2: then how you analyze the data. But before I go
4:19
Speaker 2: there and more of a philosophical point, I mean, I
4:24
Speaker 2: think that this area has traditionally been so poorly performing
4:31
Speaker 2: because of one man, Any Decards, Any Dedcard, the French
4:37
Speaker 2: mathematician and philosopher who came up with the idea of
4:41
Speaker 2: dualism and and that mind and body.
4:44
Speaker 1: Are separate entities.
4:46
Speaker 2: And I actually think you know that that that really
4:50
Speaker 2: drove a lot of Western science, and there's there's a
4:52
Speaker 2: lot of good that comes out of that from breaking
4:55
Speaker 2: things down to its smallest component parts to understand it.
4:58
Speaker 2: But when it comes to men to health, it kind
5:01
Speaker 2: of assumes that mind and brain are almost separate things,
5:05
Speaker 2: which is just completely ridiculous whenever you understand.
5:09
Speaker 1: How we operate, right, Yeah, I mean, so if you
5:12
Speaker 1: take the full architecture purely from philosophic point of view,
5:15
Speaker 1: you know is you know, if you go back to
5:17
Speaker 1: that part, the philosophical architecture of dividing the mind and
5:21
Speaker 1: body and the deductive logic around understanding it is what
5:25
Speaker 1: has led to understanding body more than mine. That's number wise,
5:29
Speaker 1: and then within body you really go narrow and narrow
5:33
Speaker 1: and narrow, and hence you know, you tend to solve
5:36
Speaker 1: a super specialized care and you are never able to
5:39
Speaker 1: think it through from the whole perspective. Okay, And that
5:43
Speaker 1: has led to the current healthcare architecture where the architecture
5:47
Speaker 1: is more for once you fall sick, then healthcare sector
5:51
Speaker 1: gets activated. Okay. Yes, and that is where if you
5:54
Speaker 1: had to redesign everything, you will not have separation between
5:58
Speaker 1: mind and body. I would even say, you will not
6:00
Speaker 1: have separation between mind and brain.
6:02
Speaker 2: Okay, yes, yes, yeah, yeah, that's that's that's another rabbit
6:06
Speaker 2: hole for us to go jarding and mean exactly.
6:08
Speaker 1: And the idea is that that separation is where the
6:11
Speaker 1: neuropsychiatry differs from neurodegeneration. So all of these can be corrected.
6:16
Speaker 1: And I would say, I do look for for look
6:20
Speaker 1: forward for the world in which we understand both mind
6:23
Speaker 1: and body together and we don't have this deductive way
6:26
Speaker 1: of solving it. And if I had to take a
6:29
Speaker 1: step further, is how do I cater to all of
6:32
Speaker 1: this a priori before the disease happens. So focus on
6:37
Speaker 1: health span and how do we enhance that. And that's
6:40
Speaker 1: your log you know, the whole logic behind longevity, that
6:43
Speaker 1: is emerging space.
6:45
Speaker 2: Yeah, and and and the other thing that you you
6:48
Speaker 2: mentioned there is prevention versus treatment. I mean, and when
6:52
Speaker 2: people talk about the healthcare system, I actually think it's
6:56
Speaker 2: poorly defined. I think we should be talking about the
6:59
Speaker 2: sick curve system, not the healthcare system. Is that because
7:03
Speaker 2: that that traditionally has been what's all about. It's about
7:08
Speaker 2: diagnosing disease and deficiency or whatever it may be. It's
7:13
Speaker 2: not about the optimization of health. So I mean, where
7:20
Speaker 2: do you find is the is the opportunity there? Is
7:25
Speaker 2: it in both healthcare and sicker or are you more
7:29
Speaker 2: interested in in prevention rather than cure or is this
7:33
Speaker 2: a twin focus of your company?
7:36
Speaker 1: The focus of the company is how to build a
7:39
Speaker 1: utility architecture of the data by which most of the
7:42
Speaker 1: decisions get made by data. Okay, how do I use
7:46
Speaker 1: the data as an architecture either to cater for an
7:50
Speaker 1: extreme patient acute you know, sick patient, or can you
7:54
Speaker 1: really build an architecture around preventing it? And how do
7:57
Speaker 1: you engage earlier in the patient so that the patient
8:00
Speaker 1: doesn't become very acute, and you'd be surprised that a
8:04
Speaker 1: typical diabetic patient by the time the onset of diabetes
8:08
Speaker 1: happens and before the healthcare system gets activated, that's minimum
8:12
Speaker 1: around ten years of wait time. Okay, So I said
8:16
Speaker 1: that again for a let's say a typical twenty three
8:20
Speaker 1: year old person gets the onset of insulin problem HP
8:26
Speaker 1: and C coming close to five point three. Okay, full
8:30
Speaker 1: lifestyle changes of prevention can really manage to reverse it
8:34
Speaker 1: for him, or if it is engaged in the earlier stage,
8:37
Speaker 1: that will be considered preventive. By the time healthcare system
8:41
Speaker 1: truly gets activated is minimum ten years delayed. And it
8:45
Speaker 1: happens when he is become insulin dependent in the sense
8:49
Speaker 1: that hpience he has gone to nine point six to ten.
8:52
Speaker 1: At that particular time, he's fully dependent on or his
8:55
Speaker 1: insulin is completely dysregular and he's fully dependent on either
8:58
Speaker 1: dialysis or about to loots. So diabetes itself has region
9:03
Speaker 1: to another disease pathology of another disease, So the come
9:06
Speaker 1: already has started. And then at that particular time, reversal
9:10
Speaker 1: is almost impossible. The containment is the only thing that
9:13
Speaker 1: is possible. Yes, Yes, our healthcare system works fantastic when
9:17
Speaker 1: it happens like you have become anuty stream to the
9:21
Speaker 1: healthcare system. Okay, if I were used to finance jobs
9:28
Speaker 1: when it comes to mental health, that ten years becomes
9:31
Speaker 1: close to thirty years.
9:32
Speaker 2: Thirty years is this before they get that get effective
9:38
Speaker 2: treatment exactly or any treatment of any nature.
9:43
Speaker 1: And that is by the time it's you know, the
9:45
Speaker 1: person has become a complete casualty in the system. And
9:48
Speaker 1: that is shocking you bad? Yeah, that is shockingly bad.
9:53
Speaker 2: And so the whole idea I'm getting a whole musk
9:59
Speaker 2: is to provide a rich data, real world data to
10:05
Speaker 2: clinicians to be able to make decisions that are effective
10:10
Speaker 2: and other data driven rather than opinion driven or this
10:14
Speaker 2: is the way that we've always done something driven.
10:16
Speaker 1: Right exactly exactly. So the simplest hypothesis is how do
10:22
Speaker 1: I learn from each patient to make the next patient better?
10:25
Speaker 1: What is Okay, a typical doctor during the training season
10:30
Speaker 1: around five thousand patients. By the time it starts practicing
10:33
Speaker 1: and gets to the peak of it as seen maximum
10:35
Speaker 1: fifty thousand patients. All his model of treatment is going
10:39
Speaker 1: to be function of either the research that he or
10:41
Speaker 1: she reads or the patient that he seen. Okay, my
10:47
Speaker 1: idea is, how do I really extrapolate and make these
10:51
Speaker 1: millions of patients so that we can learn from each
10:54
Speaker 1: one of them? Because that is how learning happens, all
10:57
Speaker 1: in a sudden that you're not taking the best of
11:00
Speaker 1: a single doctor, but you're taking the best of the
11:02
Speaker 1: industry in the sector. Okay, So right now we have
11:06
Speaker 1: data for forty one million mental health pasions. Imagine, never
11:11
Speaker 1: in the history of mankind anyone has put together this
11:14
Speaker 1: must of data at one place by which we can learn. Wow.
11:19
Speaker 2: And so to talk us through this, I find this
11:22
Speaker 2: pretty fascinating, and I like, I like the methodology that
11:26
Speaker 2: you're using. It's like, let's get as much data as
11:29
Speaker 2: we possibly can and then interrogate that data and make
11:32
Speaker 2: data certain decisions right across different people, different age groups.
11:39
Speaker 2: How do you how do you go about starting to
11:43
Speaker 2: collect that data?
11:45
Speaker 1: Right? And it's like what comes first?
11:48
Speaker 2: Like, if you're going to so presumably you're having to
11:52
Speaker 2: go to some of the big healthcare companies, are insurance companies, Like, like,
11:58
Speaker 2: what's what's the pitch?
12:00
Speaker 1: Yeah, So the pitch is when we started, I came
12:04
Speaker 1: across a specialist electronic health record system. So a health
12:08
Speaker 1: record system is a system of records that was built
12:12
Speaker 1: by Duke University, and it was built over a period
12:14
Speaker 1: of twenty years, and they had data for half a
12:18
Speaker 1: million patients for mental health. Okay, So a database that
12:23
Speaker 1: was built over twenty years and had half a million
12:26
Speaker 1: patients sitting in the archive somewhere. Okay. I acquired that
12:32
Speaker 1: company from Duke University in US, and I said, okay,
12:37
Speaker 1: let me show you the power of the data. Because
12:40
Speaker 1: that data in the raw form is sitting in the archive.
12:44
Speaker 1: Extracting from archive and creating the feedback loop and showing
12:48
Speaker 1: the value is a massive data science problem and a
12:52
Speaker 1: massive technology problem. How do you normalize disparate amount of
12:57
Speaker 1: data sets and make it in a way that you
13:00
Speaker 1: and I can ask questions and I can ask questions
13:03
Speaker 1: on the back of a fingertips rather than literally putting
13:07
Speaker 1: a scientist behind it. Okay. So heavy lifting of that
13:10
Speaker 1: is called curation of the data. Yeah, gotcha. Curation is
13:15
Speaker 1: almost like a refinery. You know, if I were to
13:18
Speaker 1: give you an analogy of say gold mine, goal mine
13:21
Speaker 1: sitting in the mind is completely different. But extraction, refinement,
13:26
Speaker 1: making the piece that you want and then selling in
13:30
Speaker 1: the market is that component is hot? Okay? Yeah, okay,
13:34
Speaker 1: And you know, being in Australia, you can easily figure
13:37
Speaker 1: out the mining or refinery is a massive example by
13:40
Speaker 1: which you can see how the physical world has happened.
13:42
Speaker 1: In healthcare, that is what one of the biggest problems
13:46
Speaker 1: is all the data in the current healthcare architecture sits
13:49
Speaker 1: in archives somewhere right. The previous version of it used
13:54
Speaker 1: to be on pen and paper, digitized and that became
13:58
Speaker 1: system of records that sits in the Electronic health recD system,
14:02
Speaker 1: in the the IT system of all the healthcare system.
14:06
Speaker 1: My idea was, how do I make the data live,
14:09
Speaker 1: how do I learn from it? How do I create
14:11
Speaker 1: the feedback loop? And that learning machine is what I
14:15
Speaker 1: started building.
14:16
Speaker 2: Okay, So how did you then go from so you
14:19
Speaker 2: got your first half a million.
14:22
Speaker 1: And yeah, so half a million robot and then that
14:25
Speaker 1: took me you'd be surprised, almost five to six years,
14:30
Speaker 1: around fifty to seventy plus data scientist and approximately around
14:34
Speaker 1: fifty to seventy million dollars to build a machine. Jump
14:38
Speaker 1: by Jesus. Right, So let me take this out.
14:41
Speaker 2: I'm quite interested in this from a business perspective, So
14:46
Speaker 2: you obviously must have had a bunch of investors that
14:49
Speaker 2: you convinced about this story to say, Hey, we are
14:54
Speaker 2: going to spend a shit heap of money to acquire
14:57
Speaker 2: this database, and it's going to take us years to
15:01
Speaker 2: analyze the data. I need to hire all these behavioral
15:04
Speaker 2: scientists and it to interrogate the data to then come
15:08
Speaker 2: up with something that we can say is usable, and
15:13
Speaker 2: then and only then can we start to get some customers.
15:16
Speaker 2: Presumably that was the kind.
15:18
Speaker 1: Of roadmap I would say, I wish you are as
15:24
Speaker 1: a ten years into it. That is exactly what it was.
15:27
Speaker 1: But when I started this, I wish I was that smart.
15:30
Speaker 1: You know, I was wrong. I was wrong on every front.
15:37
Speaker 1: So what do I mean? I was right on the
15:40
Speaker 1: front that organizing the data has a lot of value.
15:43
Speaker 1: I would not have survived if I had not come
15:46
Speaker 1: across the database from Duke University, because you do need
15:49
Speaker 1: a starting point. I can't spend another twenty years to
15:52
Speaker 1: build the database before I can build this company. So
15:56
Speaker 1: the company of our nature sits on the back of that. Okay,
15:59
Speaker 1: Healthcare has spent a lot of time in digitizing their
16:02
Speaker 1: assets and now the digital form of the data. Okay,
16:05
Speaker 1: the existence of the data in the raw form was
16:08
Speaker 1: the precondition for me to start this business. I really
16:12
Speaker 1: thought when I got the half a million patients data,
16:15
Speaker 1: I thought, Okay, it will take me two to three years,
16:18
Speaker 1: maybe three to five data scientists, maybe one to three
16:21
Speaker 1: million dollars. That was my calculation in twenty sixteen. I
16:25
Speaker 1: was wrong on every front, every front. Okay, it took
16:30
Speaker 1: me close to five to seven years. It took me
16:33
Speaker 1: around fifty to seventy data scientists, and it took me
16:36
Speaker 1: close to around fifty to seventy million dollars before I
16:41
Speaker 1: could show the value of that data in a meaningful
16:45
Speaker 1: fashion to everyone in the healthcare And when I said
16:49
Speaker 1: meaningful fashion, that missed. The validation comes from two sources.
16:53
Speaker 1: Source number one is a customer paying for it, so
16:57
Speaker 1: pharmaceutical research and just what have you a healthcare vertical
17:02
Speaker 1: person page for it. And the second is scientific publication
17:07
Speaker 1: publications that you do either with academic research or do
17:10
Speaker 1: you do with any one of the stakeholders within the sector. Okay,
17:14
Speaker 1: we are now at a place where this database is
17:18
Speaker 1: the best industrial grade and scientific grade database in the world,
17:23
Speaker 1: the best and the largest, Okay, but I had to
17:27
Speaker 1: build that machine. I had to build that refinery. I
17:30
Speaker 1: had to really really build that every component of it
17:34
Speaker 1: with extreme focused on neuroscience to show the value of
17:37
Speaker 1: the data.
17:38
Speaker 2: And did you in those five seven years where you
17:44
Speaker 2: dropped fifty to seventy million?
17:47
Speaker 1: And was that? Were you doing publications and I was
17:52
Speaker 1: doing publication. I'll do multiple small tasks. So first five
17:58
Speaker 1: years I raised around ten million. I was raising money
18:01
Speaker 1: on a daily basis. Then around twenty twenty, in the
18:06
Speaker 1: fifth year, we raised around twenty million at one go
18:09
Speaker 1: in our Series A, and somewhere around seventy year we
18:12
Speaker 1: raised another seventy five million. So you know, the sum
18:15
Speaker 1: total of all three is we have raised around one
18:17
Speaker 1: hundred five million. But you know that's the level of
18:21
Speaker 1: capital that has gone into it.
18:23
Speaker 2: If I were too, yeah, that's insane, I was going
18:26
Speaker 2: to say, that's that's an insane amount of capital and
18:29
Speaker 2: you must have endured many sleepless nights and you must
18:33
Speaker 2: be a freaking good salesman to raise that amount of money.
18:38
Speaker 1: But surely, I mean, clearly, the Vision Division people bought
18:42
Speaker 1: into the Vision straight away right otherwise, and the need,
18:48
Speaker 1: the need is very Everyone knows that the need is
18:50
Speaker 1: extremely lost. Everyone knows that it is the largest unsolved problem.
18:55
Speaker 1: You know, I can even tell you now, I can
18:59
Speaker 1: put another half a billion dollar behind it, you know,
19:02
Speaker 1: and even then I will not be able to solve
19:05
Speaker 1: every component of neuroscience as a sector. Okay, it's a
19:09
Speaker 1: multi billion dollar sector if I could really extrapolate. Okay,
19:13
Speaker 1: and if you had to think around oncology around two
19:17
Speaker 1: hundred and twenty five billion dollar business, this is a
19:20
Speaker 1: solid fifty to eighty billion dollar business in the whole.
19:24
Speaker 1: If I think of pure neuroscience one.
19:26
Speaker 2: The technology, Yeah, when you look at the when you
19:30
Speaker 2: look at the.
19:31
Speaker 1: Problem drug development to yeah, every company from therapy, standard
19:36
Speaker 1: of care, payment models, everything. If I look at every
19:39
Speaker 1: component of those solution metrics, then I would say one
19:44
Speaker 1: hundred million is almost like a peanut. You know, I
19:47
Speaker 1: can easily to capture that whole market. And I don't
19:51
Speaker 1: think you're the only one. But this is a foundational asset.
19:54
Speaker 1: It's a massive foundational infrastructure asset on which the whole
19:59
Speaker 1: sector can try upon itself without any tallege.
20:02
Speaker 2: So so the next the next thing in my mind, now,
20:04
Speaker 2: while around your whole business and business model. So so
20:08
Speaker 2: number one, there is there is proving the value. But
20:12
Speaker 2: then you know, you talk about these huge databases of
20:16
Speaker 2: healthcare information, but but they tend to be very very
20:20
Speaker 2: tightly controlled. So so how do you then get access
20:25
Speaker 2: to the other databases?
20:29
Speaker 1: Is that.
20:31
Speaker 2: Is that done through government? Is that done through insurance?
20:35
Speaker 1: Companies? Today? Own the data? Who owns that data that
20:39
Speaker 1: you then access? We go after healthcare system by healthcare systems.
20:45
Speaker 2: So so you go to talk to the NHS in
20:48
Speaker 2: the UK and the most of I mean, the reality
20:53
Speaker 2: is every healthcare institution knows that curation of data is
20:58
Speaker 2: extremely expensive business, and they are not curating it and
21:02
Speaker 2: they are not using their own data.
21:05
Speaker 1: I go and curate the data for them, and post curation,
21:12
Speaker 1: I give them their copy of the data free. So literally,
21:16
Speaker 1: think of the simplest way. Let's usum tomorrow. Take the
21:20
Speaker 1: largest healthcare system in Australia. Okay, I will go and say, okay,
21:24
Speaker 1: you have data for five million patients. Take every component
21:28
Speaker 1: of the behavioral health data that you have. Let me
21:31
Speaker 1: curate it for you. Because curating it will cost you
21:34
Speaker 1: three to five million dollars. Okay, I will do it
21:37
Speaker 1: free of cost for you, and I will give you your
21:41
Speaker 1: version of the clean data to you, and I will
21:44
Speaker 1: keep a copy of anonymized data, so I don't aggregate.
21:48
Speaker 1: You keep the aggregated d identified data. Right there you go,
21:53
Speaker 1: and with another full right to do research, full right
21:57
Speaker 1: to do everything on the back of it, and hence
22:00
Speaker 1: anything that is needed to the you know, the deidentificational
22:03
Speaker 1: version is all there. They not only get their own data,
22:06
Speaker 1: but if they want to benchmark against what the healthcare
22:09
Speaker 1: system is doing, they can benchmark their data against everyone else. Right, Okay,
22:14
Speaker 1: that's that's a mega, mega advantage to the old center.
22:18
Speaker 2: And so a key part of your business is is
22:21
Speaker 2: that momentum because the more data sets, you know, if
22:25
Speaker 2: you get Australia, if you get the UK, and your
22:29
Speaker 2: value grows because the data set is getting bigger, you're
22:33
Speaker 2: then able to do all of these benchmarks then nobody has.
22:36
Speaker 1: So that's where the momentum comes from. My whole world
22:39
Speaker 1: post is to show it for a single country and
22:42
Speaker 1: a single vertical So the single country would be say
22:46
Speaker 1: currently US US data is where we have the most comprehensive,
22:50
Speaker 1: where the rights to put it on cloud is easy,
22:53
Speaker 1: and all of that once that is done. Then show
22:57
Speaker 1: the value of these data through scientific education, commercial use,
23:01
Speaker 1: case with pharmaceutical research and working with healthca systems, all
23:04
Speaker 1: of them. So we show the value to the the hospitals,
23:07
Speaker 1: We show the value of the R and D for farmers.
23:10
Speaker 1: We show the value of any stakeholders, especially from the
23:14
Speaker 1: academic science, from Oxford to Stanford to Deacon. Any one
23:18
Speaker 1: of these guys use our data to really publish the
23:20
Speaker 1: research and so literally make it available to more than
23:23
Speaker 1: one particle.
23:25
Speaker 2: So you see, you are actually then sorry value You're
23:28
Speaker 2: actually then making your data available. They aggregated the identified
23:32
Speaker 2: stuff to researchers, to pharmaceutical companies to do research on right, Okay,
23:38
Speaker 2: absolutka can so.
23:40
Speaker 1: And that is where the hard science and tech problem was.
23:43
Speaker 1: My whole goal was, how do I take the data
23:46
Speaker 1: which is sitting in the archive clean it. Curation is
23:50
Speaker 1: the dirty piece. No one have done that plumbing work,
23:53
Speaker 1: you know, I used to say for the longest time,
23:56
Speaker 1: but I'm the best plumber in town. Okay, And the
23:59
Speaker 1: plumbing is literally you know, is how to curate the
24:04
Speaker 1: data and make that once you've done it that. We
24:07
Speaker 1: have created a platform called Neuroblue and that Neroblue platform
24:10
Speaker 1: has all the data. I can give you license to
24:12
Speaker 1: that data tomorrow. You know, if you want to say, no,
24:15
Speaker 1: well I want to play with this data, I'll say okay,
24:17
Speaker 1: let me give you access to you know, to trial,
24:20
Speaker 1: play with it, you know, do whatever if you like
24:22
Speaker 1: doing the research. And then you know, if you want
24:24
Speaker 1: to use the commercial use case, there is sixty thousand
24:27
Speaker 1: dollars and you can have that database at your fingertips. Okay, right, Okay.
24:31
Speaker 1: The reality is if you wanted even one hundreds of
24:36
Speaker 1: that data before almost was created, you just spend me
24:40
Speaker 1: in one two years minimum five million dollars and even
24:43
Speaker 1: you will hardly have one hundred thousand percents data. Yeah,
24:46
Speaker 1: you know, I like the business model.
24:49
Speaker 2: So but you talk then about you guys provide real
24:53
Speaker 2: world evidence.
24:54
Speaker 1: What do you mean by real world evidence? So the
24:57
Speaker 1: real world evidence is how does a mental health patient
25:02
Speaker 1: behave in the real world. So when a patient goes
25:05
Speaker 1: to a doctor, she's let's assume that person has depression,
25:09
Speaker 1: earlier stage depression, okay, and let's assume that Okay, that
25:13
Speaker 1: person has gone to hospital number A in Australia. Okay,
25:18
Speaker 1: what that person status is what's the phenotype of that
25:21
Speaker 1: pasion doing work not doing work? You know, the marriage status.
25:26
Speaker 2: So that part I know all of this is in
25:29
Speaker 2: their health records, right, gotcha?
25:31
Speaker 1: Exactly everything in the health then what medication he or
25:34
Speaker 1: she is taking, what doses that he's taking for a
25:37
Speaker 1: period of time, what has become effective on him, him
25:40
Speaker 1: or her and things of that nature.
25:42
Speaker 2: What has happened to their metabolic health over that period
25:47
Speaker 2: of time, right, And how the drugs interacting with the
25:51
Speaker 2: metabolic health exactly?
25:53
Speaker 1: Okay? What are the desion criteria for a clinician to
25:57
Speaker 1: look at that data and say, Okay, I'm going to
25:59
Speaker 1: switch from this this does or I'm going to switch
26:02
Speaker 1: from this medication to this medication. Okay. The full understanding
26:06
Speaker 1: of the symptoms of the patients that they have gone through,
26:08
Speaker 1: a status of the assessment of the patient, and why
26:11
Speaker 1: accruination has made the decision that he or he has made, Okay, Right.
26:16
Speaker 1: That combination is where the full phenotype of the patient
26:21
Speaker 1: becomes very live, and when you do it at millions
26:24
Speaker 1: of patient data level, then it becomes extremely live, super
26:31
Speaker 1: live platform where you can do all kind of things
26:36
Speaker 1: that you can really query in the sense all unsuddin.
26:39
Speaker 1: There is no part of neuroscience that you can't answer
26:43
Speaker 1: if you have the right framing of the question. M yeah.
26:47
Speaker 2: And because because that was going to be that was
26:49
Speaker 2: my leg Snyder thinking is how do you then interrogate
26:54
Speaker 2: the system? So your is your system set up to
26:59
Speaker 2: interrogate the data or is your system just all the
27:02
Speaker 2: data there and somebody asks a particular question and then
27:06
Speaker 2: it goes in and looks at.
27:07
Speaker 1: It and us So that is where so all of
27:09
Speaker 1: these is, in pure simple terms, raw data getting normalized
27:15
Speaker 1: and coming into common data model. Because when the data
27:18
Speaker 1: is coming from millions of sources, normalizing it and putting
27:21
Speaker 1: them into a common standard is a super complex data.
27:24
Speaker 2: That's I was going to say, that's got to be
27:26
Speaker 2: really hard because there's different healthcare systems around the world
27:30
Speaker 2: that use different processes.
27:32
Speaker 1: Absolutely absolutely, so it's very hard, not hard, very hard, okay.
27:37
Speaker 1: And to make sure that I can be exposed to
27:41
Speaker 1: every scrutiny so that the people who are going to
27:44
Speaker 1: use the data, they are all mdpsds, They can trust
27:47
Speaker 1: the data, and they can verify the sources, they can
27:49
Speaker 1: do the version control. Every kind of scientific layer is
27:53
Speaker 1: built so that it is the highest quality, gold standard
27:56
Speaker 1: of the data. That perspective, that has been my sort
28:00
Speaker 1: of motto from day one that Okay, I'm going to
28:04
Speaker 1: build the best scientific grade, industrial scale data at your fingertips.
28:09
Speaker 2: That's they There must be though, there must be a
28:12
Speaker 2: number of assumptions that you have made along the way
28:15
Speaker 2: to be able to normalize that data.
28:18
Speaker 1: And everything is everything you documented. So if you look
28:22
Speaker 1: at our data dictionary, it has runs around one hundred
28:24
Speaker 1: and eighty pages. So anyone who wants to literally go
28:28
Speaker 1: through why a certain decision was made, the taxonomy of
28:32
Speaker 1: it is extremely well documented and fully well understood. And
28:36
Speaker 1: those are the things that is required. And that's what
28:38
Speaker 1: I mean by plumbing, like if truly someone wants to
28:41
Speaker 1: build the science behind it, that is the level of
28:44
Speaker 1: taxonomy has to be built. Now, once that part is done,
28:49
Speaker 1: then the question comes down to it is how do
28:51
Speaker 1: we create because extract the insight. That is where we
28:56
Speaker 1: built the neuroblue platform. So the neuroblue platform has a
28:59
Speaker 1: coding in part I mean through which I can build
29:01
Speaker 1: all the query to querry the database. Okay, right, okay, okay,
29:07
Speaker 1: gotch and then gave the right to anyone who builds
29:11
Speaker 1: the query and the outcome. Take that insight and run
29:14
Speaker 1: with it. Do whatever I feel like, Okay, because my
29:17
Speaker 1: idea is how do I make this database and the
29:20
Speaker 1: platform is able in maximum how do you democratize the data?
29:25
Speaker 1: Right there? You go? There, you go okay, And so
29:29
Speaker 1: we have done that extremely well. We have shown the
29:31
Speaker 1: value to pharmacuity company. We have shown the value to
29:34
Speaker 1: academic researchers, we have served the value to healthcare systems.
29:37
Speaker 1: We have shown the value to even now even to
29:39
Speaker 1: academic community. I have taken this database to undergraduate computer
29:44
Speaker 1: science as students and show them the listener. Do your
29:48
Speaker 1: modeling on this data and you will have a paper
29:50
Speaker 1: that is highly research grade which you'll be able to
29:53
Speaker 1: use it for any kind of you know, So this database,
29:58
Speaker 1: why if you should be in every library in the.
29:59
Speaker 2: World that that is for somebody's just finished the PhD.
30:03
Speaker 2: There is mind blowing that somebody could go and access
30:08
Speaker 2: a database that size is just bonkers right.
30:12
Speaker 1: Exactly, the sense that there is at least thousands of
30:16
Speaker 1: different PhDs sitting on the back of the statas.
30:20
Speaker 2: MM interesting, Who who are who are your biggest custom
30:25
Speaker 2: commercial customers right now? And how do you think that
30:29
Speaker 2: that's going to change over time.
30:32
Speaker 1: I'm I'm guessed, can I can? I can? I have
30:34
Speaker 1: a guest in the battle before you get into it.
30:37
Speaker 2: I would I would guess that pharmaceutical companies would would
30:41
Speaker 2: probably be a big customer if they're not already naw,
30:45
Speaker 2: there'll be a huge customer leader. I would imagine health
30:49
Speaker 2: insurance companies in the United States because of the way
30:52
Speaker 2: the system is set up, I would imagine they would
30:55
Speaker 2: be huge customers. I mean, Australia's probably a little bit
30:59
Speaker 2: different the UK. It's probably the government and then research institutions.
31:05
Speaker 1: I think that you've got multiple customers, right, right, So
31:10
Speaker 1: I went after the hardest customer, hardest customer, hardest. I
31:14
Speaker 1: look at customers in two bucket. Bucket number one who
31:17
Speaker 1: pays for it, and bucket number two who doesn't like
31:19
Speaker 1: to pay for it. Okay, the customer who likes to
31:23
Speaker 1: pay for it in that bucket. Pharmacutul companies other best Yeah, yeah, yeah.
31:28
Speaker 1: They can make.
31:29
Speaker 2: A shareholder, they can make a shedload of money that
31:31
Speaker 2: goes straight into their bottom line.
31:34
Speaker 1: That's their business model, right.
31:35
Speaker 2: Whereas governments, Yes, it's a twenty thirty year vision that
31:39
Speaker 2: you've got to sell to them.
31:40
Speaker 1: Exactly and more than governments. The healthcare system, you know,
31:46
Speaker 1: the hospital system don't like to pay for anything, and
31:49
Speaker 1: the path through which it works is very very hard.
31:54
Speaker 1: But my take was not to get detter hite. But
31:57
Speaker 1: my intent was, Okay, how do I make it commercial
32:00
Speaker 1: really viable business model by making this data possible to
32:04
Speaker 1: pharma secular sector as a first point, okay. Then went
32:07
Speaker 1: after healthcare system at a aggregated level, so NHS is
32:11
Speaker 1: a very large customer to US, okay, And then went
32:15
Speaker 1: after economic institutions for publications purposes. They are the third
32:18
Speaker 1: vertical okay. The vertical that I haven't gone after is
32:22
Speaker 1: insurance company, government's military. You know, all of those are
32:28
Speaker 1: the next sectors that will be panned out over the
32:31
Speaker 1: next five to ten years time.
32:33
Speaker 2: Okay, because other thought, particularly in the United States, that
32:37
Speaker 2: they would be the big fish, right they particularly.
32:39
Speaker 1: The medical insurance, They are very big fish. But the
32:44
Speaker 1: time and the dollar amount and the sales cycle is
32:50
Speaker 1: very very large, very high. Gotcha. Gotcha? Like if I
32:54
Speaker 1: had to go, I mean, I would go to VA,
32:56
Speaker 1: I will go to military, I will go to you know,
32:59
Speaker 1: every of the large government institutions and show the value.
33:04
Speaker 1: And certainly the insurance companies are probably the biggest. But
33:07
Speaker 1: now we are at a level where if there is
33:11
Speaker 1: a single country that comes up on the planet and says, okay,
33:14
Speaker 1: I want to contain my mental health by say thirty
33:17
Speaker 1: percent in five years time, I can sign that contract
33:20
Speaker 1: and deliver on that value for any of the planet.
33:24
Speaker 2: Wow, and who are you allied to say? What sort
33:27
Speaker 2: of countries that you're working with? I would imagine Singapore
33:30
Speaker 2: would be pretty high on the list that from what I.
33:33
Speaker 1: Know so far. Funny enough, this company was started in Singapore.
33:37
Speaker 1: I started in Singapore in twenty fifteen. I to live there, Ascia.
33:43
Speaker 1: That's where this companyer started. Right now, the market where commercially,
33:47
Speaker 1: the market we are working is a pharmacial company anywhere
33:52
Speaker 1: on the planet, okay, but you know the certainly the
33:56
Speaker 1: largest of them are in US and Europe, So you know,
33:59
Speaker 1: move to the Maru that then, or one Japanese company.
34:04
Speaker 1: Those of our customers UH and NHS as a healthcare system,
34:08
Speaker 1: these two are the two biggest marketing which we operate.
34:13
Speaker 1: We COVID. We used to operate in a different market
34:16
Speaker 1: in Southeast Asia as well as in China and Australia.
34:21
Speaker 1: We have had multiple discussions with Ramsays, with black Dog,
34:25
Speaker 1: with Deacon, but not to a level. You do see
34:32
Speaker 1: a lot of people who are super interested but don't
34:35
Speaker 1: have the funding or don't have the budget to pay
34:37
Speaker 1: for it. And after savvy around where can I spend
34:42
Speaker 1: time you know, after all you know, uh, after the
34:47
Speaker 1: lot of money that is riding on the back of
34:49
Speaker 1: what is needed to build what we are building. So yeah,
34:52
Speaker 1: super interesting.
34:53
Speaker 2: So so if we then dive into the the what
34:57
Speaker 2: the data is coming up with, I mean, can you
35:01
Speaker 2: say from from interrogating the data or from the publications,
35:07
Speaker 2: like what are some of the big biggest pain points
35:11
Speaker 2: so the low hying fruits. What's broken in the mental
35:16
Speaker 2: healthcare and neuroggeneral disease system.
35:21
Speaker 1: I would say two or three things that stand out
35:24
Speaker 1: very fast. Number one, yes, every component of mental health,
35:31
Speaker 1: the access to care leads to much later than you know.
35:37
Speaker 1: And the trial and error that goes into treatment is
35:41
Speaker 1: very hard. A typical depressent patient goes through eighteen months
35:44
Speaker 1: of trial and error or ten different medications and then
35:48
Speaker 1: converges to two medication. Wow. Wow, that's a lot of
35:55
Speaker 1: trial and error.
35:56
Speaker 2: And that's a lot of suffering for the individual because
36:00
Speaker 2: I know that all of these medications, particularly that the
36:05
Speaker 2: psychotropic medications, that they are a shit heaper side effects exactly.
36:09
Speaker 1: So I would say even if you can reduce that
36:13
Speaker 1: by ten to thirty percent, you will see a mega
36:16
Speaker 1: again to the whole system, megagain to the individual, and
36:18
Speaker 1: megagain to everything. That's number one. Yeah. Number two, mental
36:25
Speaker 1: health is our Neurosychiatry is the largest primary or secondary indication,
36:34
Speaker 1: and especially if you think not only from mortality point
36:37
Speaker 1: of view, but disability point of view. Yeah, I mean
36:41
Speaker 1: for sure disability, Yeah, yeah, yeah, you are seeing handicapping
36:46
Speaker 1: of a person as early as thirty year old and
36:50
Speaker 1: literally a burden on you know, the quality of life
36:53
Speaker 1: is destroyed so that it may he or she may
36:55
Speaker 1: or die. But the disability is to a level of
36:58
Speaker 1: it is just not for so anymore. So yeah, so
37:02
Speaker 1: co morbidity component. And number three, if I may say.
37:06
Speaker 2: It's worth while, it's worth while double clicking on that
37:08
Speaker 2: and the valves just so that people understand, I mean,
37:12
Speaker 2: I know from from the data, just if you take
37:15
Speaker 2: and the impact on employers, Depression is the one condition
37:21
Speaker 2: that causes employers the most money and everything, and mostly
37:26
Speaker 2: through presentism, through lost productivity, not just absentism.
37:32
Speaker 1: Right.
37:32
Speaker 2: And then as you said, the comorbidities that run off
37:36
Speaker 2: that is pretty huge.
37:37
Speaker 1: Right, So the traditional healthcare things in bucket of mortality.
37:42
Speaker 1: You know, do you get the disease today? Will you
37:44
Speaker 1: die tomorrow? If it does not, then you know you're safe. Okay.
37:48
Speaker 1: The idea is not to think of mortality as a
37:51
Speaker 1: key factor. Think from quality of light metrics or key
37:55
Speaker 1: factor if you really want to go double click from
37:58
Speaker 1: the employer perspective, then recently helped think in terms of
38:00
Speaker 1: presentities in person subsenti is okay, how effective you are
38:04
Speaker 1: even if you are present and even if you're employed. Okay,
38:08
Speaker 1: And that is where the amount of the problem becomes
38:12
Speaker 1: bigger and bigger. And mental health is the single largest
38:16
Speaker 1: contributed factor across all of this. Yeah, it's certainly a
38:21
Speaker 1: wiki problem, isn't it. It's a massive problem. Yeah. So
38:25
Speaker 1: number one was essentially trial, and number two is the
38:28
Speaker 1: commolity problem, and number three is how early it is
38:34
Speaker 1: setting it? So right, Yeah, we are seeing anxiety and
38:39
Speaker 1: bipolar and as early as a year, right, and.
38:43
Speaker 2: The younger I would imagine that that is now getting younger.
38:48
Speaker 2: I mean, we're seeing that in the data with younger people.
38:50
Speaker 2: And if you get a mental health condition as a teenager,
38:54
Speaker 2: you're way more likely to have it throughout your life. Right,
38:57
Speaker 2: So you must see these warning sound just flash flash
39:01
Speaker 2: flash yes.
39:02
Speaker 1: Yes, And if I were to take a extraporate further
39:07
Speaker 1: and make the impact of say social media slash AI,
39:13
Speaker 1: then virtual psychosis is a very big component that is
39:17
Speaker 1: emerging at a very fast pace. Interesting terminology, virtual psychosis.
39:22
Speaker 1: Mm hmm. Okay, so yeah, we are we are certainly
39:28
Speaker 1: sitting at a you know, I don't mean to scare
39:30
Speaker 1: away the audience with the idea is it is a
39:33
Speaker 1: super complex problem and deserves all the attention and the
39:37
Speaker 1: funding and the science and the knowledge and the treatment
39:40
Speaker 1: to really cater to if you really want to admounce
39:42
Speaker 1: the science and cater to the health care that we
39:44
Speaker 1: want to have. Yeah.
39:45
Speaker 2: Well, I mean the other the other problem that you're
39:49
Speaker 2: trying to solve is around near to generitive diseases. Yes,
39:52
Speaker 2: and then you know, Jesus, you picked another big one there.
39:55
Speaker 2: I mean, particularly you look at dementia, even just Alzheimer's dementia,
40:01
Speaker 2: which is one subcategory of dementia like that is the
40:06
Speaker 2: most expensive disease to treat, I believe because they just
40:11
Speaker 2: need once they get a diagnosis, they just need increasing
40:15
Speaker 2: amounts of treatment.
40:16
Speaker 1: For the rest of their life.
40:18
Speaker 2: It's not like there's a cure for like like there
40:21
Speaker 2: is in cancer or cardiactors.
40:23
Speaker 1: You rehab. This is just a treatment vacuum for the rest.
40:27
Speaker 4: Of your life. And listen, I signed vortex. Yes, I mean,
40:33
Speaker 4: I would say the science will emerge at it very fastest.
40:36
Speaker 4: I'm a hardcore optimist in this area. I mean, if
40:39
Speaker 4: I look at the say the history of hocology fifty
40:43
Speaker 4: years back, pediatric oncology was death. It was very hard,
40:46
Speaker 4: whereas right now the success eight is close to one
40:49
Speaker 4: hundred percent. Okay, yeah, I look at all of these
40:52
Speaker 4: as to Okay, what are the problems that can be
40:54
Speaker 4: solved in the next tent twenty years time and what
40:56
Speaker 4: are the physicality of solving some of those?
40:59
Speaker 1: I would say neurodegeneration, especially Alzymer's and dementia of the
41:03
Speaker 1: world in a full reversal is going to be hard.
41:07
Speaker 1: You will take much longer. But neuropsychiatric modulation of neurodegenerativity,
41:13
Speaker 1: so can I fix the cognition problem or share I
41:16
Speaker 1: would say yes, impulsivety, cognition, PUP, all of those can
41:21
Speaker 1: be solved, and you will see pathway to success in
41:24
Speaker 1: that area in the cups and presumably large language models.
41:30
Speaker 2: AI is all of that being a bit of a
41:33
Speaker 2: game changer for your business because a few years ago
41:37
Speaker 2: the computational power that you had at your fingertaot is
41:40
Speaker 2: so different to HI.
41:42
Speaker 1: So I would say, if I were to really extrapolate
41:47
Speaker 1: and look into it, I'll make three statement number one.
41:50
Speaker 1: Neuroplus AI is a field that you will see, okay.
41:55
Speaker 1: Neuroplus longevity is a field that you will see, okay.
41:59
Speaker 1: But the biggest component is going to be that this
42:03
Speaker 1: database is probably the best training data set that is
42:07
Speaker 1: out there to influence the sector. Okay. So it goes
42:11
Speaker 1: back to my original argument around Okay, how do I
42:15
Speaker 1: build an infrastructure asset on which the whole sector can
42:18
Speaker 1: be built? You know, from that perspective, I have built
42:20
Speaker 1: something that I'm really proud of us, and AI is
42:24
Speaker 1: a mega tailwind in our favor because there is a
42:27
Speaker 1: lot that can be built on the back of this data.
42:30
Speaker 2: Yeah, and presumably then about that the more client slash
42:36
Speaker 2: countries that you get on board, the more successful and
42:39
Speaker 2: quickly you become successful. Right, So is there in your mind,
42:45
Speaker 2: is there a tipping point and also on your runway
42:50
Speaker 2: of your business where you think, once I hit that
42:52
Speaker 2: or get that country on board, or get to this
42:54
Speaker 2: amount of people.
42:56
Speaker 1: That's when it's really going to get momentum. I would
43:00
Speaker 1: say we are as we speak, we are on the
43:03
Speaker 1: first tipping point. Okay. So my initial intent was that, okay,
43:08
Speaker 1: if we can have around six to eight million patients
43:11
Speaker 1: data and approximately around twenty five to thirty publications and
43:16
Speaker 1: a few farmer customers, I would consider that as you
43:19
Speaker 1: make our success. Okay, Currently we have let me give
43:25
Speaker 1: you some numbers. Eleven years into it, we have forty
43:28
Speaker 1: one million patients data for twenty plus years okay. We
43:33
Speaker 1: have close to eighty plus publications and about to become
43:37
Speaker 1: one hundred plus very soon. Okay. So that's a huge
43:40
Speaker 1: amount of publication of the science that has been built
43:42
Speaker 1: on the back of it. So both commercial validation and
43:45
Speaker 1: the academic validation. If I were to use the air language,
43:49
Speaker 1: we have approximately around five trillion tokens. You know, this
43:53
Speaker 1: is probably mind boggling amount of data on which any
43:56
Speaker 1: kind of air models or large language models can be trained. Okay,
44:01
Speaker 1: so the I look at all of these as the
44:06
Speaker 1: peak of a first mountain. How do I go from
44:09
Speaker 1: here to the next mountain? And what are the problems
44:12
Speaker 1: that you want to solve for that one? That is
44:14
Speaker 1: where you know is the searching for second peak is
44:17
Speaker 1: where I am right now.
44:19
Speaker 2: Okay, so you know you're clear agar what that second
44:22
Speaker 2: peak is. That's just your The second peak is literally
44:26
Speaker 2: going after neuro degenerative for the neuroscience. That's number you know,
44:29
Speaker 2: that's a very obvious extension of this peak. Yeah, my
44:33
Speaker 2: biggest peak would be a combority market. So literally, how
44:37
Speaker 2: do I build either a commority market? So how do
44:40
Speaker 2: I understand a diabetes plus depression patient better than ther diabetes?
44:43
Speaker 1: Number two? How do I build a chat zypt for
44:46
Speaker 1: mental health? Okay? How do I build a really a
44:51
Speaker 1: very solid intelligence framework with near gero hallucination is the
44:57
Speaker 1: key you know tool, so that it can be deployed
45:00
Speaker 1: a much much distributed US.
45:04
Speaker 2: I can see how your solution or your company can
45:09
Speaker 2: be very very useful in both of reserves.
45:11
Speaker 1: Because I mean, we take Alzheimer's disease.
45:13
Speaker 2: There's been so much money spent and probably wasted.
45:19
Speaker 1: On the idea that it is it.
45:22
Speaker 2: Is caused by by plaques and tangles that a lot
45:27
Speaker 2: of researchers are now going I think we've actually been
45:29
Speaker 2: bocking up the wrong tree for twenty years, and we've
45:33
Speaker 2: spent so much money, Like if there's and I know
45:36
Speaker 2: that the attitude is changing around.
45:39
Speaker 1: What drives that, But but I mean that's got to be.
45:42
Speaker 2: Just got to have huge potential, particularly for the pharmaceutical
45:46
Speaker 2: for the pharmaceutical companies, right, it's got to be more
45:48
Speaker 2: massive then.
45:50
Speaker 1: And then the mental health problem.
45:52
Speaker 2: I mean, Jesus, if you still look at the increasing
45:56
Speaker 2: size of government spending and insurance companies in the States
46:01
Speaker 2: that is is being put towards mental I mean, it's
46:04
Speaker 2: just huge, isn't it.
46:06
Speaker 1: Yeah, So if I can take the second crack of
46:08
Speaker 1: how do I solve it back, you know, recursive back
46:11
Speaker 1: to the problem that you just said, Going after neuro
46:14
Speaker 1: regenerative is a very good idea, Going after comorbit is
46:18
Speaker 1: a very good idea. Building a native healthcare company and
46:21
Speaker 1: literally equivalent of charge for mental health is a very
46:24
Speaker 1: good idea. Really working towards influencing cost of care, including
46:29
Speaker 1: behavioral health is a very good idea. Any one of
46:31
Speaker 1: these requires has a huge market potential and requires a
46:35
Speaker 1: huge amount of time in capital comming, and all of
46:39
Speaker 1: those are multiple peaks to climb on the back of it. Yeah.
46:45
Speaker 2: Absolutely, the big thing's gott to be for you where
46:49
Speaker 2: I go after, where I put my resources?
46:52
Speaker 1: Does your system? Does it currently or will it in
46:57
Speaker 1: the future?
46:59
Speaker 2: A lie theoretical interventions like like, for instance, could you
47:05
Speaker 2: take all of this data around, let's just pick one thing.
47:09
Speaker 2: Let's just pick depression, and could you using large language models,
47:14
Speaker 2: could you based on the data that you have in there,
47:17
Speaker 2: would you be able to go, Hey, what about if
47:20
Speaker 2: we got this person to increase their physical activity right
47:26
Speaker 2: and what impact would that potentially have?
47:30
Speaker 1: And then if we made.
47:32
Speaker 2: Physical activity plus certain dietary changes early in the trajectory
47:37
Speaker 2: of the disease, and how could we are what impact
47:42
Speaker 2: would that have? And are you able to do that
47:45
Speaker 2: right now or is there plans to do it?
47:49
Speaker 1: No? I think really that is what I call is
47:51
Speaker 1: the core moderate market. So right, okay, you know, like
47:56
Speaker 1: a typical dessent also depressant EOC is g by JL
48:00
Speaker 1: people ubicity but not treated for depression. If I treat
48:04
Speaker 1: for both the obesity plus the depression with that person
48:08
Speaker 1: will get better faster, both clinically as well as you know,
48:12
Speaker 1: every other component that you can think of is what
48:14
Speaker 1: I'm trying to drive, and that is essentially I classify
48:17
Speaker 1: that as go RBID problems. And I can tell you that, yes,
48:22
Speaker 1: this is one of the largest problems that we are
48:25
Speaker 1: working on. It can be done even now today, right, okay?
48:28
Speaker 2: And could you would you be able to run? And
48:34
Speaker 2: I guess I guess this depends on how much data.
48:37
Speaker 2: The more data that you have, the more successful these
48:40
Speaker 2: hypothetical interventions. Like could you get to a point where
48:43
Speaker 2: you could run a hypothetical intervention and almost too to the.
48:51
Speaker 1: Grid of actually doing it in the real world? Right,
48:56
Speaker 1: the theoretical answer is yes, okay, we can be done.
49:00
Speaker 1: It's almost like can you run a parallel arm or
49:04
Speaker 1: a synthetic trial on a set of patients and really
49:07
Speaker 1: see that what's the expected outcome of that synthetic fact? Okay? Yes?
49:12
Speaker 1: Or a pragmatic trial all of that, I would say,
49:15
Speaker 1: the simple answer is yes. The nuanced answer is for
49:21
Speaker 1: what purpose? And who is paying for it. So is
49:24
Speaker 1: that a commercial question? And under what framework in a
49:28
Speaker 1: way do you work? So if a healthcare system tomorrow
49:31
Speaker 1: wants to build that architecture, I can tell you that
49:34
Speaker 1: I will be the best partner for that healthcare architecture.
49:37
Speaker 1: So my job is to solve the utility layer of
49:43
Speaker 1: this infrastructure. And that layer is the highest quality of
49:47
Speaker 1: the science and the tech and the data. Okay, the
49:52
Speaker 1: contracting problem and the commercial layer of that is where
49:56
Speaker 1: I'm expecting the market to react to it, and I'm
49:59
Speaker 1: trying to make that right the market. The market will
50:02
Speaker 1: tell you what it once really exactly And you know,
50:07
Speaker 1: I have I have underwritten the first risk of building
50:10
Speaker 1: this database, showing the value of it. Uh, you know,
50:14
Speaker 1: but the market has to pick and choose, you know,
50:17
Speaker 1: and that that is where it's going to be the
50:19
Speaker 1: you know, in the sense we'll see we will see
50:22
Speaker 1: which problems are relevant for the society and which are
50:25
Speaker 1: those relevant sectors that they want to come up.
50:28
Speaker 2: Fascinating ship it really really is, So what what what
50:33
Speaker 2: do you see that? But what what what excites you
50:37
Speaker 2: about the potential in this space? And what do you
50:41
Speaker 2: see as the potential roadblocks or head winds? For that path.
50:47
Speaker 1: So I would say the potential of this place is
50:50
Speaker 1: near in finite. Okay. And the reason I say that
50:54
Speaker 1: is it is, in my view, one of the last
51:01
Speaker 1: super large frontier in the history of medicine and the
51:05
Speaker 1: history of healthcare. Okay, And I don't say that likely, okare.
51:10
Speaker 1: That is a very aggressive and large sector to build.
51:15
Speaker 1: My biggest fear at this stage And if I narrow
51:19
Speaker 1: it down only to mental health is will it stay
51:23
Speaker 1: as a health care problem or does it become a
51:26
Speaker 1: social care problem? What do I mean by that is
51:28
Speaker 1: social vice in the sense mental health patients are still
51:36
Speaker 1: in the jails system. Okay, wow, wow, exactly, or they
51:42
Speaker 1: are in incarceration center. Go to any top cities of
51:46
Speaker 1: the world. You have a best cancer hospital, but tell
51:50
Speaker 1: me a single best mental health hospital in the world
51:52
Speaker 1: doesn't exist. Yeah, agreed, doesn't it? Okay? So if you
51:58
Speaker 1: have money, you send your kids to the have and
52:00
Speaker 1: if you don't have money, it goes do incarciz center, jails,
52:04
Speaker 1: I InCAR sources centers, every therapeutic knowledge is currently incarcer center.
52:09
Speaker 1: It's because this sector has treat this cause I is
52:14
Speaker 1: a healthcare problem called a social care problem because I
52:17
Speaker 1: as poor man's problem. Okay, that needs to change, and
52:22
Speaker 1: that architecture is the biggest risk that I see in
52:24
Speaker 1: the center. Having said all of that, I would say
52:30
Speaker 1: I raise a lot of optimism around the brute force
52:38
Speaker 1: of the power behind AI and the brute force of
52:41
Speaker 1: the need for the market is becoming more and more relevant,
52:46
Speaker 1: and I think we will be forced to solve this
52:49
Speaker 1: problem rather than ignore the problem. Yeah, look, I think
52:53
Speaker 1: you're absolutely right.
52:54
Speaker 2: I mean, it's it's they're both the nearer to generative
52:57
Speaker 2: sies as we as we age, you know, as we
52:59
Speaker 2: live longer ner degender diseases are are almost a kneeled
53:05
Speaker 2: on by product of living longer, right, and and you
53:10
Speaker 2: know everybody knows the burgeoning mental health problem, particularly without
53:15
Speaker 2: use and the end of role on the knock on effects.
53:17
Speaker 2: So I think you've you've definitely picked the right sector,
53:21
Speaker 2: but you definitely picked a bloody hard sector to to
53:25
Speaker 2: try and overcome. But I guess that's that's that's where
53:29
Speaker 2: the magic lies. The ball this is. This has been
53:31
Speaker 2: absolutely fascinating, really really fascinating. I love the way that
53:36
Speaker 2: you you're bringing data to to to actually solve the problem,
53:40
Speaker 2: and big data and and wow, you must be excited
53:44
Speaker 2: about the development of AI because that that's just got
53:47
Speaker 2: to be It's gotta be a rocket ship under your
53:49
Speaker 2: country if used properly. Right that, I think that's the
53:52
Speaker 2: key with any I I right.
53:55
Speaker 1: Very cool the vaal.
53:57
Speaker 2: Where can people go to learn more about you? And
54:01
Speaker 2: who who should be going and looking and learning more
54:04
Speaker 2: about you?
54:06
Speaker 1: I would say Holmers dot com is you know, is
54:11
Speaker 1: the company's for our listeners? That's h O l l
54:15
Speaker 1: u esk dot com. Yes, yes, And I would say
54:20
Speaker 1: anyone who is interested in this space, and especially to
54:24
Speaker 1: people who are really solving a super complex problem with
54:29
Speaker 1: a focus on AI and any part of the healthcare problem,
54:33
Speaker 1: whether it is governments, individual entrepreneurs. I could not have
54:37
Speaker 1: more capital and more ambition coming into the sector. This
54:41
Speaker 1: is a problem worth solving, and more capital, more ambition
54:47
Speaker 1: is the only way to solve it.
54:49
Speaker 2: Absolutely. Yes, indeed, Well, there's been awesome, really fascinating. Thank
54:53
Speaker 2: you very much for your time, and enjoy the rest.
54:56
Speaker 1: Of your day. Thank you, Paul, Thank you for having me.
54:59
Speaker 1: It's a pleasure, such a great pleasure to talk to you.
55:02
Speaker 1: And you know, I'm hoping that it was useful for
55:06
Speaker 1: your audience. Absolutely, thank you, Thank you chess.
55:12
Speaker 2: H m hm
55:17
Speaker 1: M hmmm hm