How digital innovation is disrupting the life sciences value chain

Soraya Khouadri:

Hello everyone and welcome. My name is Soraya Khouadri. I'm a senior manager in the Business Tax Incentives group, and the National Life Sciences Sector leader in the Quantitative Services division at EY Canada.

The pandemic has shone a spotlight on the life sciences and health technology sectors. The incredibly rapid development and commercialization of vaccines and treatments to control the COVID-19 pandemic have made us aware of the importance of this industry, which is a highly strategic sector and a major innovation vector for our economy.

Digital innovation has been a top priority for life sciences companies in the past few years, and the pandemic has only strengthened the notion that outdated digital tools won't cut it in this new age. Staying competitive requires life sciences organizations to accelerate widespread adoption of innovative solutions. The time has come for them to develop digital capabilities, make investments that are necessary to harness the abundance of data, and connect it to the fast-evolving health care digital ecosystem, which promises to dramatically raise the quality of care.

At EY, we believe that achieving growth begins by seizing the upside of digital innovation and transformation, tapping into the connected ecosystem to engage patients and healthcare providers differently, and improving operational efficiency and patient outcome. It is in this context that I'm pleased to introduce three leading figures from the life sciences and health technology sectors, who will be sharing their unique perspectives on the latest trends in the industry, and on how developing a digital strategy drives long-term volume and growth.

Welcome, Dr. David Jaffray; Senior Vice President, Chief Technology and Digital Officer at University of Texas MD Anderson Cancer Center; Bharat Srinivasa, Co-founder and Principal at Amplitude Ventures; and Ben Massingham, Vice-President, Head of Transformation and Innovation at Novartis Canada.

Now, allow me to introduce the moderator of our panel today, Aaron Smith, Partner in Public Sector and Health Consulting at EY Canada. …. Over to you, Aaron.

Aaron Smith:

Thanks so much, Soraya. Thank you everyone for joining us in this virtual panel and thank you to our panelists for being with us today. I’m very interested to hear from all of you. David, we’ll start with you for our first question. What is your definition of digital innovation as it applies to the life sciences sector?

David Jaffray:

It’s great to be here Aaron, and I really appreciate Soraya’s introduction on the overall opportunity and what we’re seeing from a transformation perspective.

For us, the real focus is pursuing value related to the flow of data in the organization—a move to a data-centric approach in how we’re thinking about our structure, the paradigm of the supply of data and demand on that data, and really thinking through the quality aspects of data flowing in the organization.

It came in many ways from this sort of realization during the pandemic that data sources really can change the way we think about things. We saw in Johns Hopkins this continuous flow of information, the awareness of organizations and its members of the criticality of data, and the quality of data and concerns around bias. We're seeing a real effort in a lot of organizations with focusing on their data supply chain, and this is something that's coming very quickly. Everybody knows that data governance is important but how do you make it real and how you look to pull value out of that? That's one of the big shifts we're seeing in addition to demands around virtual care.

Aaron Smith:

Thanks so much, David. One of the things I'm most excited about with the panel is that we have such a diverse set of participants. Bharat, over to you. Same question: How would you define digital innovation?

Bharat Srinivasa:

Thanks Aaron. Happy to be here, and thanks to EY for the invitation…  Let me start off with who I am. I'm a Biotech VC, and my job is identifying and investing in new therapies for patients. Further, my job is to increase the probability of success for therapies going into the clinic, and I use all of the tools at our disposal in order to move therapies forward.

One of the things we’ve been seeing is that new biological methods have resulted in a deluge of data, both internally generated as well as externally available. It’s becoming difficult, if not impossible, to do therapeutics the old-fashioned way, which is to test and reiterate in vitro, in vivo in mice.

To me, digital innovation is using computational approaches as part of the drug discovery process. Either pre, in terms of target identification; during, in terms of identifying new chemistries or doing patient stratification; or after, in terms of patient selection for clinical trials. So, what digital innovation means to me is becoming a part of the drug discovery process.

Aaron Smith:

Thanks so much, Bharat. Ben, from your perspective, I’d love to hear how you define it, especially given what we've heard from David and Bharat.

Ben Massingham:

Thanks Aaron, and thanks for the opportunity to join two such interesting people on the panel. Whenever I think about what digital innovation is at the top level, it is leveraging both technology and data to deliver better healthcare.

As a company, we suddenly are interested in everything that Bharat was talking about: how you use data and artificial intelligence to be more productive from a research and scientific development perspective. We also look at our global supply chain. My job is very specifically Canada and the way that Novartis is set up. Our operations tend to start once a product is already on the market. So, whilst I'm fascinated with what Bharat does, my job doesn't do any of that sort of thing.

We really focus on two challenges in particular, and I think digital innovation needs to play a much bigger role. The first is how you engage with your customers, particularly physicians and patients. The second is how you can address together some of the big problems in the delivery of healthcare. That's not necessarily a traditional role for the pharmaceutical industry. But by building a coalition between policymakers, by building coalitions with private public partnerships, with physicians, with patient groups, with the technology ecosystem, I think we're in an age now where we can really solve some of those big problems within the delivery of care. In this case, everybody wins and 99 times out of 100, technology—and all the innovation and ecosystem that comes with it—has to be part of that.

Aaron Smith:

I think there are interesting definitions across the board. We covered almost the entire value chain, from drug discovery through patient engagement, and clinical trials. One of the key aspects is that the definition of digital innovation also evolves as you think about the value chain, and from computational power at the beginning through engagement tools at the end. Management and the maintenance of data is such a critical aspect of that, so we’re seeing large tech companies making acquisitions in the health and life sciences space.

We've just talked about the importance of technology across the board. So, do life sciences companies need to evolve to become technology companies to remain competitive? Bharat, we'll start with you on this one.

Bharat Srinivasa:

The simple answer is “no,” and let me define that in a bit more detail. Life science companies, especially in the biotech realm, which is what I'm most knowledgeable about, have a tough enough job with advancing therapies. And for you to evolve into doing something completely different from what you were built for and are hiring people to do is difficult. It's not about evolving into something else, but it's adding on capabilities internally to help you increase your probability or increase your productivity.

I mentioned the pre, during and post. A lot of companies still focus on doing experimentation over and over again. But is there a way for some of these companies to add on AI capabilities, since there are so many companies right now in the pre-discovery process, to identify new therapies or new generated chemistries that they can advance into the clinic? Can they, during the discovery process, use AI or different computational approaches to develop new pipeline programs? Can they identify new patient stratification approaches to increase the probability of success? And then post, can they do that from the clinical trial perspective? So, it's not a matter of evolving; it's more about adding these capabilities into your value chain.

Aaron Smith:

Ben, what's your perspective on that?

Ben Massingham:

Fundamentally, if I had to give a one-word answer, I would go with Bharat. No, but I suspect you probably want a bit more out of a public discussion than one-word answers. So, I'll elaborate: yes and no.

I think there are certain things that, as an industry, we need to do and own ourselves. One of the things we're focusing on is how you use digital innovation to better engage with your customers. Other industries have been doing that in-house and we should be, too…

I used to work for GE, so I know what it takes to be a great diagnostics company. That's not us. That's not us as a pharma industry. And that's where, when Bharat says “no,” I think it's through partnerships, through collaboration. We have to become tech savvy and we have to become users and disseminators of technology. Does that mean we become builders? I don't think so. Does that mean it's the core of our business? No. So, tech-enabled? Absolutely—we will not survive if we're not tech developers. …

Aaron Smith:

Pretty consistent so far. David, what's your take on this?

David Jaffray:

I'd have to agree. I think the key thing is that it's not about adopting digital technologies—it's about changing the way we work in response to this digital capability. And I see the technology as easy; the change is hard.

On our data strategy side, the cultural shift on focusing on the stewardship of data is a very challenging, massive change in the organization. We talk a lot about computational approaches to predict outcomes for patients or predict better drugs to go after. Getting people to adopt that and realize that's the way it's going, that change, from the traditional low-yield drug development paradigm to something that's higher yield. Getting that shift, the skill set shift, the recruitment, the rather-than-return to the old methods, but instead pick a “Moneyball” approach on compute and prediction instead of based on what the recruits were saying about the talent. It takes a lot to make that change and I think we all will change. But the question is those who change fast enough will lead the pack.

Ben Massingham:

I think it's a really interesting point. I spent half my time thinking about what technology or problem to solve, and half my time on the culture piece, as you say. Some people find “new” threatening, they find “new” difficult. Some people are all over it, , sometimes you have to rein them in…. I think we often underestimate the cultural challenge of doing things differently, especially in an industry like pharmaceuticals where we’ve probably had the same business model for 25 to 30 plus years, and people have been very successful with that. There's no burning platform and it's tough.

Bharat Srinivasa:

And it’s something we face in our day-to-day job as well. You're working with CEOs, and see the new breed of CEOs saying, “Yeah, let's do data.” You think, “Whoa, slow down.” And then you see the old breed of CEOs saying, “No, we're going to do 5,000 mice to make sure we see what we want to see.” And there's something in balance with that, but it's trying to thread that needle and make sure you're progressing the biology forward.

David Jaffray:

Yes, Bharat, I can imagine. I don’t know if you remember that scene in “Moneyball,” if you’ve seen the show. Billy Beane is trying to convince the scouts that the way they swing the bat is not the way to choose the lineup. I can imagine you talking to a bunch of scientists and drug development teams that have gone around several times, and convincing them that the compute suggests we go this way instead of that way. I can imagine that's a very similar conversation.

Bharat Srinivasa:

But it always starts off with a lot … when I say AI, and then it's a matter of walking them through what that means.

Aaron Smith:

I’d love to dive in on that a bit more, Bharat. In our prep, we talked about some ways that you and the firms you're investing in are disrupting some of the traditional pathways. How are you finding that acceptance to looking at more than how fast a runner runs to first, to pick a baseball team?

Bharat Srinivasa:

We've used two approaches. The one that has been most successful is going into companies early and building them out from the ground up. So it's always, “Here’s the biology strategy, and here’s the computational capability strategy,” and hiring people across both teams who can speak to each other, as Ben and David mentioned. That's been what's most successful for us in terms of advancing it forward.

On the other side, we have looked at companies that are later-stage, and there it's a matter of having someone who can speak both languages, both computational as well as biology. But that conversation is different. Then, it's a matter of putting increased weightage on one over the other as the company evolves, in order to advance that forward.

To give you an example, we invested in … a couple of years ago and they just did a big CDC round. Now, that's a company we've been looking at for the last five years, and they initially started off as, “We want to do AI and AI will solve it, and we will identify things and partner with other people.” That evolved when we did the … to saying, “We can use the AI to identify a pipeline, but now we are a pipeline company enabled by the AI.” And then, taking that to the next level with this current CDC run, which is saying, “How can we now industrialize the AI to spit out 20 different programs and raise the capital required to advance forward?”

So, it really depends on the stage-by-stage basis. But getting back to your question, the easiest way is just to build it up from the ground up. But when we do company creation, as we do at Amplitude, to build in both capabilities from the ground up.

Aaron Smith:

That's a great segue into the next topic, which we've already touched on. How should life sciences companies and the sector leverage technology and data differently to generate value? We talked to the different parts of the value chain. So, David, do you want to kick us off on this one?

David Jaffray:

I think in this case, changing the way we work in response to digital capabilities is key. We're seeing the traditional paradigm continue with the real-world evidence conversation. Companies are looking to aggregate data based on secondary-use de-identification, and some of the pharma companies are extremely interested in that topic. At the same time, a lot more understanding of the importance of the patient data and the consent framework, and the academics are understanding the value of the data.

As the value of the data becomes greater and understood, I think we're going to see a lot tighter partnering between healthcare organizations, industry partners and compute teams that work on things like synthetic cohorts to give control arms with larger and better matching. And we're seeing that go all the way to regulatory. So, what I see happening is much tighter partnerships with industry, academic and healthcare.

In a way, that's not the kind of Google-Ascension throwing data over the fence. Not that it's a bad thing, I guess. But the point is something that's much more strategic, much more intentional and much more aligned, and the technologies to allow that to happen are coming where it really tracks how data is being used, and for what use. I think it'll draw teams closely to become more efficient and will take some of the friction out of these processes. I'm very excited about that.

Aaron Smith:

What about from your perspective, Ben, on the downstream aspect of the value chain? What are you seeing needs to shift?

Ben Massingham:

I think it really comes down to what the value that we need to collect to be, to provide the system. I mentioned earlier that I think the biggest bang for your buck in terms of improving patient outcomes is not by developing the next drug. I think that will always be important and we'll always need to do that. It will be around how you just get the right patient at the right time to the right doctor to make the right choice.

There are so many examples of people who end up at the emergency room who never needed to with the right preventative care. Whenever I think about all the data that exists out there, and David was talking about that sort of potential for collaboration, if you can not only aggregate and access that data, but if you can also interrogate it proactively to identify patients who can benefit or are at risk of not benefitting, I think that's the sort of value you can provide.

Now, there's data that exists, but there's also data that can generate. One of the things that we're looking at, just as an example, is depending on different diseases, the disease progresses, and sometimes you can see it and sometimes you can't. Multiple sclerosis is one of those areas where it's much harder to see the progression. It's a horrible disease, but it tends to happen over a prolonged period.

Technology can help you generate the data you need to be able to tell where that patient's at. So, it's not just about interrogating existing databases. It's what technology allows us to do to create new types of data. Without going into too much detail, a company called Innodem we've entered into a partnership with can track progression of the disease through eye movements on an iPad, and we're looking to validate that. But that's amazing because then you're not only saying this person isn't doing as well as they should, that we should think about doing them differently, but you can also imagine the next step where a physician is proactively reminded and said, “You need to see this patient.” Now, you don't need to wait for your appointment in 18 months. And similarly, this patient is doing really well, so we probably don't need to see them.

It starts to get into healthcare utilization as well as better care. So, I think if we can combine existing data and new technologies to generate new data around genuine actionable patient insights that can help the system manage those patients better, well beyond just choosing the drug, then I think everyone will win and ultimately the patient will get better care, but at the right time.

Aaron Smith:

David, are you seeing the realization of that, knowing you've moved from the system here to a new system in the US, like where the leveraging of some of these technologies may be different? How could that apply to Canada?

David Jaffray:

I think it applies universally. The engagement around data is differential between the US and Canada. The HIPAA frameworks in the US provide very clear secondary use on the identification. That's not so clear in Canada, but the concepts are still there. And the European paradigm is actually pushing substantially from a GDPR perspective in terms of how we will end up using this kind of data moving forward. I think they have the high ground and eventually we'll see that being adopted more broadly.

It's in the interest of the individual, and hopefully society will see that direction, then we'll get some consent frameworks that allow us to do that. So, I think it's going that way. I think there's always opportunity like a jurisdiction, like Canada, to take a leadership position in this. If we could just get our provinces to agree, it would be fantastic not to be fragmented. Small population fragmented is a smaller population. It would be nice if we could get some opportunities to level that out, but I totally agree with Ben's comments.

What we're finding is that when we've talked to our industry partners, or we're trying to help advance our mission against cancer, they're interested in advancing our mission against cancer. So, we're collaborating around data to make that happen. The next thing that happens, once you bring those two parties together with a focus on the quality of data, is what should we measure next? What are we missing? What clues could we use to better personalize the intervention? What cohorts are the ones that Bharat's computations suggest we should be getting an effect from? So, that all starts to become quite connected, which we call a supply chain. There's a good demand and understanding of where it's coming from and you might want more of that, or we want something else. That's a very powerful feedback loop.

Aaron Smith:

Bharat, from your perspective… how are you leveraging those downstream databases to model out the behaviours of the molecules and products that your companies are developing?

Bharat Srinivasa:

There may be a different answer to the question, which is that the market for existing data is extremely crowded. If data exists already, it's already cited across multiple different hospitals, and quality needs to be measured. And there are not 1,001 companies going after the same data. So, what we've been trying to do is be a little different… For a biotech company that's doing CDSAD runs, it's hard to compete with a Novartis or a GSK on access to that data, as well as have the bandwidth in order to intricate that data.

Part of what we're doing from that approach is working with AI discovery companies who already have the reach to get access to the data. But even that's starting to become crowded because you can only go to itemized … so many times before you start to see the same analysis come up. In our space, there are new techniques that are developed that can analyze biology data better. We just invested in a company that spent about $100 million in about three years on going out and collecting patient samples from about 30 different hospitals, and using that to build out their own internal library of single-cell sequencing data. That's something that only they have access to; nobody else does, so we have a good line of sight into where the data came from, and as you would say, what the provenance of the data is.

You have the right questions to ask in order to credit the data, and then feed that into the patient stratification tool going forward. That's given them a leg up, and it also takes us away from this extremely crowded, competitive space of existing data and moves into the less crowded space of generating new data with all of the quality metrics that you would assume account for. That's why I mentioned the AI plus biology strategy. We need to find people who can speak that language, and we've been looking for that exclusively.

Aaron Smith:

Ben, you mentioned that one of the ways of working differently is through partnerships. You referenced one. Is that something you're seeing shift more and more? We talked about if this group agrees it was unlikely that pharma-biotech companies will become technology companies. How are you seeing that partnership space evolve?

Ben Massingham:

I see more and more of these partnerships coming into being. I think it's a little early to see how they actually work and what they do. I can certainly comment on what we're trying to do. I think when you say you want to be a partner it needs to go beyond fund. We've done that historically, through grants and donations or through partnerships, but there are a couple of different areas that those partnerships can go.

One is the facilitator of a conversation and creating focus around the needs. We've just got off a big event focused on cardiovascular health and what was really interesting is that we were able to bring together policymakers, technology companies and pharmacists, and we learned a couple of things about that. One, people are surprised the pharmaceutical companies want to have that conversation with them, and you have to behave in a certain way. One practical thing is that we put all these people around the table, and then everyone from Novartis, apart from one person, left because you need to position yourself as an equal partner around a joint agenda, not as the owner of the agenda and I’m going to push you in the direction I want. That’s how those partnerships will need to evolve just in the way of thinking, but it comes down to mutual value. If you can rally around a mutually valuable proposition or problem that you all have a vested interest in sharing, you can work like that. If someone’s vested interest is bigger, it becomes problematic.

The second area, and the pharmaceutical industry is renowned for this probably with good reason, is that we like a good process. We have regulatory things that are important, but as a result of that, we can be sometimes be a little hard to work with. One of the things we’re trying to do is make it a lot easier for startup companies with a really interesting potential piece of tech to do pilots with us, and that’s all under the banner of the bio. If you’re interested, there’s a whole raft of things written about it on the Internet. But if you go through the traditional procurement safety, all of these processes, you will literally kill these startups. It’ll take them six months to fill in all the paperwork, and in those six months, they’re not earning any money and they’re not developing anything. They won’t survive. We're realizing that if you want to get integrated into the ecosystem, you can bring value by focusing on them. You can bring value by investing in them. You can bring value by really helping them come to market. We have commercial experience that a lot of these start-up companies don't. But you can kill them by being too -overbearing and too process-driven, so we're trying to get that balance right.

David Jaffray:

That's a great point, Ben. We've been working with some platform technologies that we should allow industry partners and us to work more closely together. The friction is really painful. I just love the way that the dial starts to make us think differently about how we work internally to bring the quality of the data along.

To Bharat's point, without the providence on the data, no one's going to trust anything in the future. That's pretty clear. You can have synthetic basis; you can do synthetic everything. You'll have so much fake medical data floating around, and I'm sure the FDA stresses about this all the time. We want to know where it came from. What's the quality? How do you prove its providence before it goes anywhere? And this is going to be a big part of the partnership moving forward.

The market for data is just finally maturing. Market forces are just finally coming into space there, and the quality metrics and providence tracking. That's where the partnerships are going to be key because people are going want to partner with people who know that they have robust data. To your point earlier, Aaron, is there opportunity in Canada? I think the government should think substantially about investing in the quality of data that's coming out of the healthcare system. It's a huge asset, but currently there's no funding for data quality; not in the way that industry is going to need it to go forward.

Aaron Smith:

Interesting points, David and Ben. Moving on, what are three ways in which you think we can continue to improve the innovation ecosystem in Canada's life sciences sector? Bharat, we'll start with you.

Bharat Srinivasa:

Wearing my Biotech VC hat, I fully agree with what David and Ben have been saying. You need to have a good system set up in order to have providence, and it’s hard to advance these programs forward.

But, gun to the head, three things: the first one is just more cross-pollination. What we've seen across the Canadian ecosystem is that you have Mila in Montreal, AltaML in Alberta, you have Vector in Toronto; you have something else there. As with everything Canadian, you see a very provincial focus on building out AI capabilities. And those need at some point to merge amongst themselves and along with the existing bio ecosystem in order to get the cross-pollination to advance good companies and good biology. That creates bigger companies here, which increases the chance of success in more countries.

The second one is to simplify the tech transfer process. There's a lot of data and a lot of technologies that exist in universities. But when trying to get them out, there are multiple, different silos, and you just start to lose time and effort with trying to bring these companies out and merge them with the broader cross-pollination. Simplifying that approach and making it easier to start companies that have this broader AI computational approach built into the ground, as I mentioned earlier, makes it easier as well.

Finally, from the company's perspective, although it's starting to change a little bit, I think we still see this notion amongst entrepreneurs here of, “I need to take this one drug, do one experiment, and then flip it to the highest bidder.” I think that's just involved and that's changed completely. There needs to be a mindset change as well in saying we need to build large companies, and we need to build out and add on all of the capabilities and tools we need in order to make this company successful. And to do that at scale is what we need across the local ecosystem.

Aaron Smith:

Thanks, Bharat. Ben, same question to you.

Ben Massingham:

I liked the point you made, Baharat, about the provincial system...

I think there are roles we need to play. The idea about focusing on a big problem and talking problems to solve, rather than technology to get excited about, is something that we can do. I think we can co-market more and bring things to the market. We have a very large footprint.

The other part of my job, and I think this is a responsibility we all have, is Canadian innovation for Canadians and Canadian industry. But then Novartis, and all other pharmaceutical companies, generally speaking, are a global organization. Part of my job is to make Canadian innovation available to the world. We have a whole series of processes. But my ambition and part of my job description is that if they have the same problem in France or in the UK that we're experiencing in Canada, I need to make sure that not only are my colleagues in those countries aware of the fact that we have a potential solution, but also put them in touch and start the collaboration so we can grow businesses through global reach and global outreach, as well as just from a Canadian perspective. I think that's a core part of what value we can bring to the Canadian ecosystem.

Aaron Smith:

David, I’d love your perspective as our Canadian down south right now.

David Jaffray:

I've commented on a few things already, but I think there's a huge opportunity for us to really leverage some of the data assets across the country. We did an interesting project with Terry Fox in the cancer domain called Digital Health and Discovery Platform, which is a distributed approach that lets people keep control of their own data and, of course, institutions, but then has a framework for a brokerage on increasing access to that data. That thinking is something that needs to be leveraged more. I said funded and supported that; I think they're making some progress now.

We're doing something similar here as an initiative called Breakthrough Cancer, which funds MD Anderson, Memorial Sloan Kettering, Johns Hopkins, MIT and Dana Farber, all working together on cancer. We've been working with them as a data governance framework that will facilitate low-friction interactions between those organizations to leverage data, from the point of view of accelerating progress against cancer. It seems like there's an opportunity.

The investment in AI in Canada, to Bharat’s points, is fantastic. It's training a lot of people to be recruited in the United States right now. We need to have more places there to be recruited into Canada and have a greater stickiness. That stickiness could be thought of in a lot of different ways: royalties, frameworks, innovation and startups. I don't think we're fully capturing what's being built there. The trainees are fantastic and the talent is fantastic. This is the follow-through of the next step of making sure they stay. To Brad's point, I think this is key; and that means to Ben's points—connecting more across the country.

Aaron Smith:

You've all brought up some of the challenges around connecting across the country. On the flip side, have you seen good or promising examples of this between provinces or jurisdictions? They're getting some of that information sharing, data sharing and collaboration, right?

David Jaffray:

As I mentioned earlier with the Terry Fox initiative, this really got started earlier in the year in many ways, but I think that concept is robust. There's a long way to go. I thought it was a good idea because I'm convinced that the way we currently do MTAs and … use agreements has got way too many lawyers involved. Half of that could be automated technology to support that standardized patient of consent. It just creates a really powerful environment and that could be leveraged moving forward, so I think more in that direction is key. Jurisdictions that do that will be attractive to industry, and sticky.

Aaron Smith:

Thanks David.

Ben Massingham:

It's a telling question… The examples of ones where I've seen it work have really been between institutions that are motivated by a research where that works around specific diseases. Within the research communities, there's often some great collaboration, data sharing and things like that. At the provincial level, in my experience, it's not always consistent within the province, let alone across provinces. In my experience, the one that's probably farthest down the line seems to be Alberta; they've got a really good data set there. They're investing in it and there's a ministry behind that.

You've got some other areas… but it's much more fragmented by virtue of the IT infrastructure and the data capture infrastructure… and I doubt that Canada's any different from any other countries in the world with so many providers of data storage and data run analytics. I'm sure there's fragmentation across the globe.

David Jaffray:

We are seeing some good opportunities from the way we interact with industry data. We refer to them as alliances. Where the partnership's much broader, the allocation of resources is dynamic. We're pursuing things together. It's cooperative teams and scientists working in both parties to further advance under a broader envelope, and that gives an agility for investments within that scope. We find those to be really powerful.

Then some internal work put together; some biotech-type pipelines internally that stay close to the academic setting and close to the clinical context. It’s stripped-down pipelines internally that can accelerate certain things forward. There's an opportunity there, and to Bharat's point, maybe we could think about how the computation could facilitate that. Those are always a risk, but what happens in those settings is that someone gets on their hobby horse and they ride it into the ground; basically their hopeful target. But computational approaches could de-risk that. Maybe we could see more embedded pipelines in healthcare ecosystems that could benefit from a lowering of the barrier there, but also benefit from the computational side to de-risk and take away bias in the decision-making. I think that dynamic is very interesting.

Bharat Srinivasa:

I feel like we've been very negative here; there are some positives as well. In terms of trying to get the government to change, Ben spent a couple of minutes talking about the friction points with big pharma. Dealing with big pharma is a cakewalk, when they talk about dealing with governments. So, I think it puts things in perspective.

One of the things that we have seen, and it's something I brought up before in terms of the three things required, is building bigger companies here. Industry does a fantastic job of getting across provincial borders and international borders and pulling up data, and that's the good. The bad, if you want to call it bad, is that it's still heavily focused on a specific area. Some of the things that we have been trying to do internally at Amplitude is determine how we work with our entrepreneurs, and how we work with both existing-portfolio, current-portfolio companies that have been successful in leveraging data for specific programs across the ecosystem, in order to scale that up. That's something that had some movement because it's very specific. But then it goes back to the point that if we are looking for that data, there are 25 other companies looking for that data as well, so the price starts to grow. It's not something that you can solve easily. There are discussions moving, but they're moving at the pace that you would expect the government to move. So, it's a longer-term solution.

Aaron Smith:

We could have a whole other webinar on how to better access and leverage public private partnerships with government, and it's an evolution for sure. So, it's not necessarily a negative, Bharat, but it can be a challenge. Last question, then we'll turn it over to some of the questions we've had through the chat. We'll start with you on this one, Ben. How can we attract the right talent to the life sciences sector in Canada and what does that talent look like?

Ben Massingham:

Fundamentally, what we love is that you have to talk differently about what the job is. You have to think differently about what the job is. One of the things we underestimate is that the mission of the health care industry is a tremendously powerful motivator for a lot of people who are not in it today. So, if you want to bring people from the outside, that's it. We have to change the way we talk about mission.

If you'll indulge me for two minutes, a couple of years ago, I posted for a brand manager I absolutely didn't want. I wanted someone from outside of the pharma industry. I wasn't allowed for very appropriate legal discrimination reasons to say, “If you have pharma experience, please don't bother applying on the job spec.” But that's what I essentially wanted and we put it up on LinkedIn, and I had three people apply who were all pharmaceutical brand managers. I didn't know why.

We have a very talented person in the organization, who knocked on my door and very bravely said, “If you're open to feedback, I know who you want. I know that they're in my network, but with that job description the way you've written it, I will never forward it on to them because it harms my personal equity.” So, once I got over the shock of being so totally useless at work in job descriptions, I challenged her. I said, “Okay, well help me write it,” and she rewrote it with me. It used words that I as a seasoned pharmaceutical executive would never have written, but we got the people. So, it taught me that how you talk to people who are different from you is critical.

The other part of this, and it's what I worry more about, is how you retain people once you get them into the pharma industry in particular. We tend to be quite slow, we're risk averse, we like a good process. And a lot of these people are prepared to make that jump from something they know very well to something they don't know at all; they want change, want dynamism, want new, want exciting. So, we need to forge them a career path that is beyond the traditional what you must carry the bag to be a brand director or something else, which is what we like doing. But we also need to try and get things out of their way and allow fast moving. So, attracting is hard, and retaining may even be harder as we go down that.

Aaron Smith:

Bharat, from your perspective?

Bharat Srinivasa:

I think it's everything Ben said, and I would add compensation to that as well. There does exist the right talent, especially in the digital innovation space within Canada. You see Mila, you see Vector spitting out fantastic graduate students, day in and day out. The problem, though, is that each of these individuals, before they even graduate, have a half-million-dollar job offer from Google, or from whatever company. And it would be stupid not to take it. So the question is, how do you—absolutely as Ben said—attract and retain them? But also, how do you pay the appropriate compensation for these individuals to build that up? And that feeds into a whole other discussion about how they build their companies and how they fund these companies, but compensation is critical.

Aaron Smith:

David, having been out of the market in Canada for a couple of years, what's your perspective on what we can do better here in Canada to attract the right talent?

David Jaffray:

I think there's a huge opportunity. The work-from-anywhere paradigm means that Canadians can stay in Canada and work almost anywhere on the planet eventually. They should do everything they can to allow Canadians to work remotely. Keep the revenues coming in and keep the talent. I think that's fantastic. We're trying to figure out how to support access from all of the planet to work with us. A part of it is the data collaboration architecture. How do we get people in to know and touch the data they can touch and the work that they're doing with us? If you build that fabric, the talent in Canada will stay in Canada, pay taxes, build equity. I think it's a huge opportunity.

How do you become the strongest AI machine learning workforce in the world, working from Canada bringing equity home? I know that we would sign up a number of people right away. Unfortunately, there are restrictions, like you have to live in the state of Texas to be employed by the state of Texas. So, there are some rules like that, but that's how rapidly it’s becoming an issue across the entire country, the United States, so any kind of remote work barrier issues have to be explored and leveraged. Huge opportunities there.

We're also seeing a huge increase in the cost of data science talent. It's pretty scary. At the same time, people are much more open to risk. They're willing to try stuff to explore things. I think that's been a very interesting part of the pandemic, a shift in the work, changing jobs, the ability to tolerate change. Before the pandemic, everybody thought you had to follow a certain path and do certain things. And now it's like, “Well, I can live through that, so we can make a change.” That from my mind is a hugely positive opportunity moving forward, and shakes things up a bit and accelerates change.

Aaron Smith:

Working for MD Anderson from the cottages in Muskoka sounds like a compelling proposition for some.

David Jaffray:

Yeah, not bad, eh?

Aaron Smith:

We're supposed to be bringing talent the other way, but it goes both ways. I think we're going to see that flexibility materialize in a real way. So, a few questions from the audience. That's great, people are engaged. I think we had over 240 people sign up to attend today. This first one is for you, Bharat. The question is: “How difficult is it to find people that speak both languages within the companies that you drive?” So, on both the biotech and the AI side of things, are you finding that you're training people in biotech on AI, or the other way around? Or are you finding there are some people with skillsets across both areas?

Bharat Srinivasa:

That's the hardest to find, honestly—someone who speaks both languages, both bio as well as AI. The way we say it, and I think that Ben said it amazingly, is that the work we do in the healthcare space is not just something you do for a job, but it also has impact on patient's lives. There are certain people with the phenotype that have exactly that: they want to do both bio, they want to do both AI. So, we are looking for them, they do exist. But going back to the competition point, Google can pay three times the salary. So, it's one of those, but how do you identify them? You can, but then how do you read in and come to them to make sure they're doing that? Deep Genomics is a great example of a company that's done that and has hired multiple people.

Aaron Smith:

Ben, here’s a question for you from the audience. What is the impact is of digitization on value-based contracting? So, the ability to track patient outcomes between the companies and linking that to both payment and if they ask about the investor's ability to monetize that.

Ben Massingham:

It's not exactly my area of expertise, but it's certainly something that we've been talking about as an industry for a long time. I can't think of many other industries that get away with saying, “This is the price, whether you get an outcome or not.” You buy a car and it doesn't work. You take it back to the garage and ask, “What the hell's wrong with my car?” We, by the very nature of what we do, have a product that works differently for different people. So, this idea of pay for performance or pay for outcomes is not new.

I think it comes down to a lot of what David was talking about, about accessibility of the data. Where we've typically found this, where these become difficult conversations, is deciding if you agree on what great outcomes look like. You form an agreement of what is an acceptable level of performance “yes or no” for your product.

The second step is, do you agree on how that will be measured and measured objectively? And then, do you agree on how that will be captured, analyzed and paid? I think there are starting to be some examples within the industry of where this is happening. But if we go back to that “tick-tick,” the Canadian example of being able to manage that across the different provinces… through PCPA negotiations, through the different tech data requirements, it's going to be difficult. So, if someone can solve for it, I know the pharma industry mostly would say we're in because ultimately you want to be able to provide value. And where it's not valuable, not only do we all believe that the patient should no longer be on the drug, but it's okay if we don't get paid for that drug … to be using it. But making it a reality, that's very different and it's getting closer, but I still think we're quite a long way away from it from a data accessibility and an interpretation perspective, particularly in Canada, but everywhere.

Aaron Smith:

David, you worked in a hospital here that spent a lot of money on drugs, and there’s a hospital there that probably spends even more. What's your perspective on that, given the vast investment?

David Jaffray:

The value-based care paradigm is active in all jurisdictions. There's no question that there's pressure. There is a lot of work going on, so the quality measures across hospitals now is quite remarkable. Platforms… across the United States, looking at performance issues and then also connect against supply chain side on the same thing, so there's a growing sector in building out how you measure value and how you facilitate access to product.

I think it's going to be really interesting to see how the industry fits into that equation moving forward. I know when we were in Canada, we're talking about directly buying product and guaranteed performance through large government contracts. If you don't want to squeeze the hospitals out, that makes it very difficult for them to be in that incentive chain. Very frustrating. So, whatever you do, you have to make sure you align incentives all the way across and the incentive has to relate to patient outcomes. I think this is where there's a really interesting opportunity to get much more patient engagement in the outcome measure. And I think things like AI, machine learning, bots and interaction can be huge there. We can talk about that, but there's a lot of technology that could be developed there to start to measure burden and response, and to validate some of the outcomes data that we believe we're getting with some of those interventions. I think that's a huge trend we didn't talk much on today.

Ben Massingham:

You're right, David, it's a really interesting conversation. But also the time horizon, as you move into more preventative care, what you're investing in versus what you're getting in the long term in terms of economic value. So, not in terms of outcomes and avoidance of something in 20 years. It just becomes increasingly complicated, but you're right: it's coming and I would actually welcome it, but how do you make it happen?

David Jaffray:

I think that the computation side is huge, and going back to Bharat's comments, we can predict what the interventions might look like in populations. Right now, we're doing it for global warming or climate change, demanding an intervention way before the impact at some level. So, the paradigm’s there. Getting the measurements and then predicting the impact on the populations, the intervention, when we have to learn to trust the predictions to act sooner. I think that a very powerful feedback loop between data and compute is going to be activated.

Aaron Smith:

One of the questions is how we know we can trust the systems that will be put in place to help inform the development of products, the security of information and compliance with regulatory requirements. Is there something that needs to change or something that needs to be more focused on?

David Jaffray:

I don't want to be negative, but it is pretty remarkable that we've produced stuff today. If you look at the chain of control data and assets, and I think in general, there's a huge lift coming on: greater standardization and measurement, quote data. It feeds into regulatory, it helps us learn to use AI and machine learning algorithms in places to make sure we're not using them in a biased fashion against inappropriate inputs. All that is a pulling up of the socks on the data supply chain. That's why I'm a big fan of that topic, because it will give back, and we'll give back on Bharat's equation, we'll give back on Ben's equation, give back on government costs and patient experience. It's just stuff we’ve got to do. It's a data economy. We need the thinking of the structure that banks brought to the financial economies that start to let the data economy come to life.

Ben Massingham:

I agree, David. I think data privacy is rightfully a hot topic and we need to be concerned about that. But if anything, I find we get in our own way of providing value through being too conservative and too many things, and we work in an industry that is heavily regulated for a very good reason. We have to be respectful of that, but I feel like sometimes it's not because it’s not Novartis because of the industry. It takes us several months to make a decision because we're worrying about all of these things and then you make it and then it can go.  I think we need to do a better job of communicating all of those steps we take to make sure that the data is appropriately preserved, rather than putting even more steps in.

David Jaffray:

Normally, I would agree with you. People get nervous when there's ambiguity and a lot of the technologies we use to move into are loaded with opportunities for ambiguity. So, getting what I refer to as data governance technology is built out to make it clear and explicit, for what you use and what's the result. And then building the consent framework on there, remarkably more structural, I think will release more data. And we'll see things starting to go in that direction.

Aaron Smith:

Bharat, any final thoughts on that topic?

Bharat Srinivasa:

I'm either fortunate or unfortunate enough that I haven't have to deal with that. But we did have to do it with one company that we invested in, and it took them a year to sign up a hospital. These are VC-funded companies with only limited ability and time, so if we can speed things up, it helps us as a company. It helps as a VC, but more importantly, it helps patients because in the end, that's what we're trying to get. We're trying to get better therapy solutions.

Aaron Smith:

I think we learned a lot today. I know I did. We started off with innovation across the value chain. We spent a lot of time talking about data and access, and then privacy and security. Those are the things that are underpinning innovation. Innovation happens when you have predictable processes, you have data you can leverage, you understand the innovation sandbox, and then you bring ideas to it. So, I appreciate our panelists for taking the time today and providing their perspectives. I appreciate the EY team that supported this and set this up, and I appreciate all the participants on Zoom who joined us. Thanks so much, and we'll chat with you all again.