What is going to be the impact of AI near term and then obviously a longer term impact that you foresee?

The amount of personalization that we can do for our customers and for our employees using AI is going to be significant.

I think we shouldn't be buying and selling AI, but we should be buying and selling AI readiness.

We have some of the biggest voices in this industry already cautioning about an AI bubble and things like that.

There are a lot of commoditized areas and I think as we get into more and more personalized medicines, drug discovery and with all the pressures that we were talking about, those things can have a significant impact of AI.

We are living in a world of unprecedented changes. We have lived through an era of change while we saw the industrial revolution and the internet and digital eras. However, the pace of change in AI is not something that we as humanity have ever dealt with.

It presents us with fantastic opportunities, but at the same time, it can be very scary. Hello everyone, I'm Subhro Mallik and I'll be your host for this series on Talking Breakthroughs. In this series, we will talk to industry stalwarts, movers and shakers and thought leaders to learn from them how each one of them have embarked on this journey to unravel the potential of this unprecedented change that we are seeing in this industry.

Welcome to the first session of Talking Breakthroughs. We are going to talk today about the amazing world of pharma and life sciences where we are in exciting times where we believe we will pivot to a whole new world, especially with AI coming into the forefront. It is my privilege to welcome Saurabh to this very first session of Talking Breakthroughs.

Saurabh is the president at HFS. Welcome Saurabh to the session.

Yeah, thanks Subhro for having me here.

Saurabh, if you look at our industry, the life sciences industry today, which both you and me are pretty privileged to have been a part of, if I may say, we are living in a different world because I think if you think about what is happening in the industry, we are seeing a huge amount of revenue pressures, a huge amount of cost pressures across the bulk of the industry, whether you take pharma or even medical devices, we are seeing a huge amount of cost pressures and people are looking for newer and newer ways by how to address both their top line and their bottom line challenges. So from your vantage point, given the intense amount of research you have been doing personally and at HFS, what would be your views on the state of industry today?

I think you're absolutely right, Shubhro. I think this is unprecedented, the amount of change that we are seeing.

And I think it goes a little bit beyond revenue and cost pressures. Because there are three or four factors that are driving this in my mind, Shubhro. First of all, you look at the geopolitics, you look at the Big Beautiful Bill, you look at Gov.rx, you look at CMS starting to negotiate pricing.

Yes, that drives cost pressures, that drives revenue pressures, but it takes away the ability for pricing freedom from pharma, which was a big, big ability and a capability for them is the freedom to price. With that gone, you know, that will have some butterfly effect beyond revenue and cost pressures that sort of starts to change the business model itself. You look at CMS starting to negotiate, you know, drug discovery and R&D.

And I think the 15 to 20 year time period that we had as a pharma industry will now get reduced to maybe 10 years. So 5 to 10 years gone, that's going to have dramatic impacts beyond revenue and bottom line that has a strategic impact on how we think about this industry. Look at the patent cliff that we are on, I think 180-200 billion is what we can foresee.

Maybe there are others, there are more biosimilars, there are more generic drugs, every payer wants to go after them because they're lower cost. So the brand is getting diluted. You look at the big breakthrough drugs, you know, those things are probably things of the past.

We are going to see more and more personalized medicine, more and more AI related things. And I think the boundaries of the industry, what we used to call pharma is also changing, you know, pharma and medtech are no longer two different industries, they're sort of converging, I don't know where, which line starts to get blurred, because most of the new medicines will have a combination of them. And you know, the way that we are doing drug discovery is also starting to change.

So I think it's unprecedented, and I completely agree with you, it's revenue, cost pressures. But I think the how we define this industry is almost changing.

That's a very powerful insight, I agree with you, the lines between pharma medical devices is blurring.

I think medical pharma companies are trying to become medical devices company and the other way around too possibly is happening. And you've touched upon an interesting topic, which we cannot not talk about, which is AI. What does it mean for us? It is a topic that you want to get to very quickly.

There is a huge amount of hope on this AI that is underpinning, not only I think, our industry, but also broadly speaking, every industry. So if you look at that as a paradigm shift that's happening, you know, what is your views as you look at it? You know, what is going to be the impact of AI near term, and then obviously, a longer term impact that you foresee that the industry can either be worried about or look at it as an opportunity and address some of their core issues?

Absolutely, we cannot have this conversation without talking about AI, Subhro. I think first of all, AI is not a flavor of the season.

I think AI, agentic AI, gen AI, these things are going to change how we work and how we operate, how different industries operate, whether it's life sciences or other industries. But I think hope is not a strategy, right? And I think right now we are, there's a lot of AI washing going on in the market, both from tech providers, as well as enterprises and everybody in the industry. And I think one of those things is the amount of obsession that we have with productivity when it comes to AI.

The first thing that, you know, whenever somebody talks about AI, the first thing is what's the productivity? What's the impact on headcount? What's the impact on operational efficiency? And I think we are doing ourselves injustice, if we are just talking about productivity. Because at HFS, I think the impact of AI, we call it the four P's of impact. Productivity is one of them.

Absolutely, it has to be there. But I think there are three other P's. One is personalization.

The amount of personalization that we can do for our customers and for our employees using AI is going to be significant. You know, we can have no call waiting times, it's going to be zero call waiting time because of the amount of personalization. The other P is prediction, you know, the ability to make smarter, better, faster, quicker decisions, whether it's strategic or operational.

That's the power of AI that we can start to take. We're starting to see that happening in supply chain, in operations, in finance, and several other areas. And I think performance, you know, how your business performs, what's the impact on business outcomes.

We will realize the value of AI, or at least the ambitions or the hope that we have with AI, if we stop looking at it just through the narrow lens of productivity, and start to look at this broader, almost like a balanced scorecard of these four P's. Otherwise, I just fear that, you know, we will create very aggressive business cases and we will not land up there.

No, very aptly said, I think, today, we look at it at a very narrow lens, there are very, very few organizations who are thinking in terms of performance, actually, with Infosys, we are also breaking it into two parts, we call it client.ai and services.ai, services is about the productivity, how can we get more efficient, client.ai is about how can we bring better business value beyond just the productivity part of it.

So I think the four P's is a good way to look at this whole paradigm and where do we fit each one of the pieces within the things that we are doing. But if you look at it, you know, the pharma industry and actually more so the medical devices possibly going ahead, has been typically laggards in adopting technology, right? But it has, at least for me personally, it's been a bit of a pleasant surprise to see how fast the industry is trying to adopt AI, and not only in the, within the P, the first P of what you mentioned, but also possibly the other three P's. So you know, what is it that you, from your vantage point, you're looking at it from an adoption, but what is it that you're seeing in the pharma industry? What's your views? And second, you know, as we think about this paradigm shift, as it applies to pharma, it's a highly regulated industry, what should leaders, what should the leaders and managers think about as they navigate this change that's coming?

Yeah. So I think, I think you're right that pharma, medtech, medical devices have been laggard, but I would qualify that laggard term that they've been a laggard in SG&A. I think if you look at drug discovery, if you look at R&D, if you look at that, and they've been a leader in the space of even using AI and other technologies, robotics and everything else. So they've been a leader in certain aspects of, which is core to their business.

But they've probably lagged, you're right, in, you know, when it comes to IT operations, SG&A operations, and those kind of things. So I think you're absolutely right, on that front. And I think what we've seen is not just in pharma, but practically every industry, everybody has, you know, a pilot or a POC in every nook and corner of their organization when it comes to agentic AI or gen AI, but hardly 5 to 7% of them actually scale up.

So I call it death by 1000 POCs that's happening right now, Shubhro. And I think that's what we need to get over with, because if pilots are very easy to prove today, you know, the finding a use case is not that hard there are, you know, every person that you talk to will have 10 examples of how AI can make their lives better. The technology is almost there as well, you know, firms like Infosys and others are, you know, I think there are enough nerds out there to get the technology there.

So I'm not worried about it. I think it's the how, how do you how do you adopt this thing is the is the challenge. And then that leads to the second question is what should people do is get AI ready.

I think we shouldn't be buying and selling AI, but we should be buying and selling AI readiness. You know, data, for example, our research indicates that for every dollar that you spend on AI, you probably need to spend $3 on data, your data is fragmented. Not that people don't have data, but data is just not ready to be used by AI.

Process debt is another huge issue. You know, it's are you happy just being slightly better, slightly cheaper, slightly faster? Or do you have to fundamentally redesign your processes? Remember the RPA days, Subhro, you know, that was the problem. It wasn't that RPA was a bad technology, but it was a bandaid to extend the life of legacy.

You know, we will go through the same thing, you know, AI is not some magic wand. You will have to do process engineering, people, you know, you look at, I think our research indicated 45% of the employees are actually scared of the AI initiatives. Now, if that's happening, then how are you going to democratize AI if people are scared of it? So, I think and tech debt, right? If you look at Pharma or even other global 2000 large companies, there were thousands of these spaghetti of systems.

And now you introduce even a small thing, there's going to be a butterfly effect to that. So, I think Pharma leaders should not just jump into becoming you know, this AI first, because not everybody can be AI first, somebody will have to be AI second. But they need to be AI ready first.

And I think that readiness is not there right now. Until you're ready, you know, whether it's AI or whether it's some other sci-fi, it's not going to be, it's not going to sort of drive our ambitions to realization.

That's a nice point.

Actually, we don't have to be AI first. I think AI first is already done.

Yeah, AI first is done and somebody needs to be AI second.

And I think, look, Netflix wasn't the first streaming platform, right? But it's the biggest and the best. Now, Google wasn't the first search engine, you know. So, I think this obsession with first is something that's great for marketing.

But is it great for value realization? I'm not too sure.

And I think we have a very apt point you have mentioned. The whole point about value realization has to be front and center to our conversations with anywhere, right? And we at Infosys also try to make that when we advise our clients, we're trying to make it very easy to spin up an NDP, like you said, but very difficult to industrialize it.

Now, we have to take industrialization and that's where I think today. So, if I look at what's happening today, right, a lot of people are realizing and some of them are actually pulling back a little bit who went into this AI first journey, trying to be the first, realizing that there is enough dollars spent without a reciprocal return that they can see or make it real. So, if you think of the industry today, do you, in your mind, as you think of it, are there some use cases that stick out to say, hey, look, this is something that you saw with some in the vast set of clients that you deal with.

And, you know, this is the kind of use case that industry needs to think about and could be pharma, could be outside. And because we do think, we do believe that the brakes are being hit on this initiative. We have some of the biggest voices in this industry already cautioning about an AI bubble and things like that, and potentially there will be a bubble.

But still, within this noise that we see, I'm sure there was certain either a use case or use cases that when you look back and say, wow, that's amazing. So, what are your thoughts on that?

I think I'll take my own personal example, Shubhro. I started my career as a software engineer at Infosys.

So, I'm an Infosys alumni in that way. And what did I do as a young software engineer? I did three or four things. I did as is documentation, right? I would say this is what the client is doing.

I'll say to be documentation. This is what it will look like. Then I would code and then I would test.

And we had thousands of Saurabhs doing this work at Infosys. Do we need that now in today's world, AI across the software development lifecycle? You know, that can almost be done, you know, if the amount of coding that AI can do for you is phenomenal. So, I think SDLC is, I think, the prime case.

Whether you do it at your horizontal IT, whether you do it for OT, whether you do it for pharma stuff, it's all the same. So, I think SDLC is, I think, the first thing that we should attack. Then there are other things, you know, customer care.

Everybody has a bad experience with customer care. And there was this study that was done which said AI is actually more empathetic than humans are when it comes to talking to people. Nobody likes to wait for 30 minutes to get somebody to answer a simple question.

I think that's the level of personalization that I was talking about. And it's possible today. It's already happening.

You know, you look at simple things like accounts payable and payroll. Why do you need thousands of accounts payable to just, you know, transact invoices? That should be done, man. Like in today's world, why do we, that is not what humans were born for, doing that kind of work.

And then I think those are some very low hanging fruits. You know, there are a lot of commoditized areas and we can go on, Subhro, that I think we have to get there. But then there are, you know, what you mentioned value creating or client.ai kind of initiatives.

Look at drug discovery. Look at how fast we came up with the COVID vaccines. That wasn't possible without AI.

And I think as we get into more and more personalized medicines, drug discovery, and with all the pressures that we were talking about, you know, those things can, they'll have a significant impact of AI. Science, molecular science, supply chains, you know, the supply chains are being disrupted left, right and center because of geopolitics and tariffs right now. Creating a digital twin of your entire supply chain and, you know, saying if this goes wrong, then what happens is possible today with AI.

It could have taken, you know, a team of 20 analysts spending data crunching, you know, over two weeks to come up with an answer. And by that time, it would have already changed. Yeah, so I think there are tremendous amounts of use cases.

I think there are use cases, there are a diamond dozen use cases that we are seeing successfully deployed. The question is how many of them will scale up?

I think Saurabh, you are right. Some of the low hanging fruit and possibly, yes, they're all worried about AI.

Maybe it will give us a greater purpose as, you know, the mundane tasks go away. And my view is, yes, there may be some initial pain for sure. Because of the impact of AI is going to have on people, we will have to deal with those nuances.

But over the longer period, hopefully, we all come out much stronger. So that brings me to an important question that we should all ask ourselves is if you had a crystal ball, you know, and you had to project 20, let's just say 30, 40 of the year, what's your view? You know, what do you think the world will look like? More importantly, in the pharma context, I have to say, you know, if you had to think about pharma in 2040, what would you like it to look like?

Oh, man. I think AI that we've been talking about is a rising one.

It's right here right now. I think this is a two or three year or maybe a four year conversation. I think the next space that we'll be talking about is physical AI, Subhro or bots that can think, you know, robots that can think because right now robots can do stuff, but that humans think on behalf of them.

But imagine a world where robots can think, you know, and with the world of physical AI and imagine things like autonomous labs. You know, that's possible where there are these thinking robots who are mixing enzymes and creating molecules and coming up with personalized things for Saurabh's conditions or Subhro's conditions and that's not possible today.

So, I think physical AI is the next big horizon that I'm looking at. And then probably a third one that I'm super excited about is space. Because AI is going to need a lot of data centers and data centers are going to need a lot of energy.

But imagine if you have data centers in space. Space is pretty cold, right? So, you don't need all that energy consumption. To create that and there are some very teeny pilots going on in that area.

But I think we will have to solve for this energy crisis that we are on the verge of as we start to adopt more and more AI. And I think those are the things that excite me. Now in between throwing quantum somewhere, Subhro, because especially for the pharma industry, I think quantum computing is going to have tremendous impacts.

Our traditional classical computing cannot solve for many problems. It can't even, we can't think of where we've come but we cannot figure out the molecular structure of a coffee bean today with classical computing. So, there will be a lot of need for quantum computing or new ways of computing as well.

But that's where at least my head is. I think it's going to be very, very exciting. I think we just need to keep learning.

Well, that's fantastic thoughts, Saurabh. I hope next time maybe by 2040, we'll be interviewing a robot and he'll be more empathetic and we will see how that goes. Now, thank you for being on the discussion today.

It's been wonderful talking to you, Saurabh. And thank you all for joining me for this session on Talking Breakthroughs.