>>Welcome to this cube conversation that is a part of the AWS startup showcase. I'm lisa martin I've got with me now the ceo of amazon john Vitaly john welcome to the cube >>lisa. It's a pleasure to be here. Nice to see you. >>Likewise give our audience in a real liaison and what it is that you guys do specifically in pharma and life sciences. >>Well, you can find that in our, the name of the company is on uh, we think of us as leading uh, customers to the horizon of AI and pharmaceutical, biological manufacturing. And uh, we're all about helping our customers take The step into Pharma 40 and really realized the value of leveraging, machine learning and artificial intelligence in the manufacturing process so they can get higher yields and predictability and ultimately better outcomes for their patients. >>Is your technology built on AWS? >>Absolutely. From the ground up. We leverage, yeah, we leveraged as much as we can from AWS innovation and, you know, a few years ago, when our founders envisioned the future of manufacturing in this industry and where it needs to go first thought was go with a leader to build the solutions and of course A W. S. Is by far the largest provider of this type of technology. And we're happy to say that we're helping and partnering with A W. S. Two to advance the science of artificial intelligence in life sciences. And uh it's just a natural fit for us to continue to leverage the platform on behalf of our customers. >>I like that. The Ai horizon. Excellent. So talk to me a little bit about, you know, the last year has been presented many challenges and also opportunities for people in every industry. I'm just wondering what are some of the changes that we've seen? Farm and life sciences companies have become household names for example, but talk to me about some of the the key initiatives in smart manufacturing and what pharma companies require. >>Well sure, you know farmer companies and biotech companies like look into the lessons from other industries where ai has been widely adopted. If you look at uh manufacturing and other industries has been widely adopted for a number of years. Tesla is a great example of how to use A. I. And robotics and and data science uh to advance uh the efficiency of manufacturing globally. Uh that's exactly what we're trying to achieve here in in life sciences. So um you know, a lot of the leading innovators in this space have been working in their labs with data science teams to you know find new ways to collect data uh to cleanse that data, make it data that's useful across the enterprise. Um but they haven't really tackled, you know, continuous processing in manufacturing yet. There are a number of leaders that are mapping out strategies and they've begun to go down this path. Um But most are really looking at how first to bring the data together in a way that it could be democratized and anonymous in some cases and used across the enterprise. Uh There's a model that we've adopted in terms of our product strategy and how we engage customers and that's the uh the the pharmaceutical maturity model which was developed by the bio forum. This maturity models is a great way for companies and vendors alike innovators to look at how to help Advance their capabilities from one level to the next. And so we help customers understand where they are in that journey and we look for the areas where they can get traction more quickly. They can see value sooner and therefore the adoption would would be accelerating across across their their sites. And in different ways of use. >>Is that maturity model? That farm of maturity model? Is it is it built on or based on digital transformation? >>Absolutely. It's all about digital transformation. And so the model really begins with pre digital and you'd be amazed to find I think the the amount of Excel spreadsheets that are still used in manufacturing today and that would be what we would consider to be pretty much pre digital because that data is not accessible. It's only used by the operator or the user. So it's really about getting from that level to uh breaking down data silos and bringing that data together and harmonizing the data and making it useful. The next level would be about the connected plant actually connecting machines and data lakes um to begin to get more value and find find more ways to improve the processes. And then you move up to using advanced analytics and AI and then ultimately have an enterprise wide adaptive manufacturing capabilities, which is really the ultimate vision, ultimate goal. Every manufacturer has. >>One of the things that we've been talking about for the last 14 plus months or so is really the acceleration in cloud adoption, digital transformation as really a survival mechanism that many industries undertook. And we saw all of us go remote or many of us and be dependent on cloud based collaboration tools. For example, I'm curious in the pharmaceutical industry again, as I said, you know, we we know that the big three and for household names that many of us have been following for the last 14 months or so. What have you seen in terms of acceleration? Informal companies going all right, we need to figure out where we are in this maturity model. We need to be able to accelerate, you know, drug discovery, be able to get access to data. Has that accelerated in the Covid era? >>Covid has been the great catalyst of all time for this industry. Ah and I think it was a wake up call for a lot of, a lot of people in the industry to recognize that uh, just because we have the highest quality standards and we have highest level of compliance requirements and um, we ultimately all think about efficacy and patient safety as our goal to achieve the highest levels of quality. Everyone agrees with that. What the realization was is that we do not have the capacity in any, any geography or with any company, um, to meet the demands that we're seeing today demands to get product to market the demand to get the supply chain right and make it work for manufacturing. The, uh, the uh, The opportunity to partner to get there was, you know, you can see that by the way companies came together to partner for COVID-19 vaccine manufacturing production. And so, um, it was a wake up call that it's time to get over the kind of cultural barriers, risk aversion and really come together to coalesce around a a smart manufacturing strategy that has to be combined with a G XP or good manufacturing compliance standards. And that has to be designed in to the technology and manufacturing processes Together. That's Pharma 4.0, >>got it. Thank you. Let's dig in more to that GSP compliance. And you guys, we talk about that in different industries. The X being, you know, X for X type of industry, talk to me about the compliance regulations and your G XP AI platform and how you guys built on top of amazon, help customers evolve their maturity and facilitate complaints. >>Absolutely. So as I alluded to earlier, one of the biggest challenges is just getting the data together in a place that you can actually manage it. And because there's so many legacy systems and on predominantly on prem technologies and use today, cloud is starting to gain a lot more traction, but it's been limited to uh kind of tier two and tier three data. Uh so now we're seeing uh you know, the recognition that uh just having a data link isn't enough. And so uh we have to overcome, you know, the biggest barrier is really a version to change and change management is really a huge part of any customer being successful. And I think with a W S and us, we were working together to help customers customers understand the type of change management that's required. It's not enough to say, well, we're going to apply the old techniques and processes and use new technology. It just doesn't work that way. If you're adding people uh, and scaling up people just to do validation, worked on a brand new platform, like AWS offers, like we offer on top of AWS, you just won't get three return on investment, you won't get the outcomes and results you're targeting. Uh you have to really have a full strategy in place. Um but you can, and start in small ways, you can start to get traction with use cases that might not have the a huge impact that you're looking for, but it's a way to get started. And uh, the AWS platform is, you know, a great way to look at um, a strategy to scale manufacturing not just in one site but across multiple sites because it's really a data management strategy uh for us using US components uh to build our data collection technology was the starting point. So how do you bring this day together and make it easy and with low overhead and begin to use Ai at the point of collection? So we built our technology with AWS components to do that it's called we call them be data feeders and those are agents that go out and collect that data and bring it together. We also because of the way at AWS innovated around data management we can use a multitude of components to continue to build capabilities on top of what we have today. So we're excited to partner to follow the AWS Roadmap but also continue to add value to what A. W. S. Does today for customers. >>Right? Seems very symbiotic but also your gives you the platform gives you the agility and flexibility that you need to turn things on a dime. I like how you said Covid was a catalyst. I've been saying that for a year now there are things that it has catalyzed for the good and one of those that we've seen repeatedly is that the need for real time data access in many industries like life sciences and pharma is no longer a nice to have but it's incredibly challenging to get real time access to high quality data. Be able to run analytics on that you know, identify where the supply chain in the manufacturing process. For example things can be optimized. Give me an example or some examples of some of the use cases that you guys are working with customers on. I imagine things like that to process optimization, anomaly detection. But what are some of those key use cases in which you really excel? >>Well, it all starts with with what we can do around predictions. There's a lot of data science work being done today, understand variability and how to reduce deviations and how to get more um of predictions to know what is expected to happen. Uh But a lot of that doesn't get applied to the processes. It's not applied as a change the process because that requires revalidation of that entire process. Our platform brings huge value to customers and partners because we do the qualification and validation on the platform in real time. And so that eliminates the needs to go back out and deploy people and uh track and re document uh and re validate what's going on in the process. So that that just takes a huge uh responsibility in some cases liabilities off off of the operators and uh the folks analyzing the data. So that's that's really to get to real time. You have to think carefully about how to apply apply ai because a I was developed in a scientific way but you also have to apply it in a scientific way to to these critical processes in manufacturing. And so that's that's only done uh on a platform, you can't do it on a kind of a stand alone basis. You have to leverage a platform because you're analysing changes to the data and to the code being used to collect and analyze the data that all has to be documented. And that's that's done by our capabilities are using to audit or create audit trails uh to any changes that are happening in the process. And so that's a critical critical process monitoring capability. That is almost impossible to do manually. Uh Some some would say it's impossible to do manually. Uh so uh the the ability to to qualify algorithms to validate in real time enables real time manufacturing and there's a F. D A. Uh I would I would say mandate but guidance called continuous process verification cPV that they will be coming out with additional guidance on that this year. That's really there to uh tell tell manufacturers that they should be getting to real time capabilities. They should be driving their investments and and types of deployments to get to real time manufacturing. That's the only way you can predict deviations and predict anomalies and deal with them in the process and track it. >>So give me give me a snapshot of a customer or two that you've worked with in the last year as they were rapidly evolving and adjusting to the changes going on. How did you help some of these customers extract more value from their pharma manufacturing processes, understand what it is that they need to do to embrace A. I. And get to that real time. >>Absolutely. So, you know, most of our customers are facing the challenge and dilemma that just adding more people and more resources and even upgrading existing technologies or adding more data scientist has a limit. They've reached the limit of improvement that they can make to these processes in the output in manufacturing. So the next natural step would be to say, okay, what science can I apply here and what technology is available To really get to that next one or two improvement in the processes. And it's really critical to look at um you know, not just one use case, but how can I address multiple problems using the same technology? So bringing multi variant uh multi variable excuse me. Um analysis capabilities um is is something that's done in every other industry um but it has not been applied here in terms of changing how manufacturing works today. We can do that, we can we can do multi variable analysis in real time, we can predict what will happen. We can actually alert the operator to make changes to the process based on uh a number of predictions of what will happen in a batch or series of matches in manufacturing. We also bring unstructured data into those calculations that wasn't possible before cloud technology came along and before a I was deployed. Um So now we can look at environmental inputs, we can look at um upstream data that can be used for improving um you know, the yield on batches. So the you know, the main um focus today is you know, how do I get, reduce my risk around asset management? How can I improve visibility into the supply chain? How can I reduce deviations in these processes? How can I get more yield? How can I optimize the yield uh in any given batch uh to improve uh you know, the entire process but also reduce costs in each step of the way. Uh So uh the good news is that when you apply our technology and our know how uh there's an immediate positive impact. There's a customer, we're working with very large customer where we walked in and they said we have this problem, we've reached a certain level of optimization and yield. We can't seem to get it to go any higher. and within six weeks we had a solution in place and we are saving them tens of millions of dollars in material loss just in that once one step in the process that's worth hundreds of millions of dollars in terms of finished product. Uh and if you apply that across multiple lines and across multiple manufacturing sites for that customer, we're talking hundreds of millions of dollars of savings, um >>significant impact, significant business impact that your customers I saw on the website, you know, R. O. I. And was at six when I get this right. I had it here somewhere um quite quickly. But the key thing there is that these organizations actually are really moving their business forward. You just gave some great examples of how you can do that. And just kind of a phase one of the project. Let me ask you this in in a post Covid world, assuming we'll get there hopefully soon. Where is in your opinion? Um Ai and ml for pharma companies, is it going to be something that is is for those that adopt it and adopt all the change management needed to do that? Is it going to be kind of the factor in deciding the winners and the losers of tomorrow? Okay, >>well, I don't want to lay down predictions like that, but I would, what I would say is uh all of thought leaders out there have have openly shared and privately shared that this is exactly where the industry has to go to meet the demands. Not just of ramping up COVID-19 vaccine production on a global basis, which we have to do. It's also dealing with how do we how do we uh scale up for personalized medicine, which requires small, small batch manufacturing? How do we turn over lines of manufacturing more efficiently to get more drugs to market more different types of drugs to market, how to contract manufacturers deal with all these pressures, um, and still serve their customers and innovate. Uh, there's also the rise of generics there, you know, that's bringing on cost pressures for big pharma particularly. And so these are all moving the industry in the right direction to respond to these on an individual basis. Would would definitely require the use of Ai and Ml But when you bring it all together, there's a huge huge of push for finding and finding breakthroughs to increase capacity and quality at the same time. >>Yeah, tremendous opportunity. My last question for you, john is a bit more on the personal side. I know you're a serial entrepreneur. What drew you to a zon when you have the opportunity? I can only imagine based on some of the things that you've said. But what was it that you said? This is my next great >>opportunity. That's a great question because I asked myself that question, uh so having been in the industry for for a long time, having been with very innovative companies my whole career, uh I knew that uh manufacturing had fallen behind even further in terms of innovating using the latest cloud technologies and ai in particular, I knew that from running another company uh that focused on the use of predictive analytics. And so uh given all the vectors coming together, the market pressure that's happening on the technology, absolutely. Being a maturity level that we could we could make these things a reality for customers in the size of the challenge. And market opportunity was just overwhelming. It was it was enough to make me jump in with both feet. So I'm very happy uh to be leading such a great team and amazing, amazing talent at amazon and super excited about our partnership with a W. S and where that's going and solving very, very complex and very critical, uh, challenges that our customers are facing together as partners. >>Absolutely. Well, john, thank you for joining me today and talking to us about who is on is what you're doing, particularly in pharma and life sciences, smart manufacturing and what you're enabling in a covid catalysis sort of way. We appreciate you joining us here today. >>This has been a pleasure. Thanks for having me. >>Likewise for john Vitaly, I'm lisa martin, you're watching the cube.