 Welcome back to this CUBE special presentation here on the 62nd floor in the Mandalay Bay in Las Vegas for Google Cloud Next. I'm John Furrier, host of theCUBE. We're here with a program accelerating innovation with Persistent. We've got two great guests here. I do John Gerhard, who's the Vice President of Cloud Technology and Product at DeepEalth. Thanks for coming on, appreciate you. Thank you. And Kamal Puri, but Cloud runs Google Cloud at Persistent. You're an AVP. Thank you, John. Thanks for coming on. This is a customer segment. DeepEalth, you've got the keys to the kingdom. You've got the cloud. You've got the product and technology. You get a lot of responsibility. We're here at Google Cloud, and the announcements are pretty impressive. It's for major change over to this transformation with the cloud. It's a huge opportunity. You guys in the middle of it. Take a minute to explain what you guys do, and then we'll get into the use case. Yeah, I think so. Before, let's say, we get into technology aspects. I want to give a little bit of insight into what we do at DeepEalth. DeepEalth is a fully-owned subsidiary of a company, RedNet. So I think so. RedNet, maybe a lot of people are aware of what it is, but it's still, like, a couple of seconds. RedNet is one of the largest diagnostic imaging chain in the United States. And we do over 10 million scans across the United States. And one of the vision from our CEO, Dr. Berger, was how can we make the healthcare accessible to people and also use technology as a means to drive health and wellness across the populations what we see. And principally, we are focused on imaging, but also we work with a lot of hospitals where we do JV with them to run their imaging business. So in a sense, like, say, we are like co-partners with them in the industry. But in a sense, the RedNet sole focus is imaging as a core part. And DeepEalth is a technology arm of RedNet, what we call informatics business. So I think so a few years back, Dr. Berger had this vision saying that we need to control our destiny in technology to drive value to our customers. So we had a lot of informatics assets, what we acquired. We acquired Pax assets. We acquired the radiology information system to the EMR essentially for radiology business. And also we had acquired few AI assets in the area of MAMO, area of lung and then prostate cancer, right? So these were all the assets within the RedNet. But our CEO, Dr. Berger, saw that the vision where if, let's say, technology could be an arm which is actually helping our core business imaging, which we can actually add more value to our customers and also improve care to patients. So recently, like say, a year back, they had the vision of creating an informatics arm of RedNet. So now essentially I think so a couple of months back, we announced to our public market that DeepEalth would be a fully separate business unit having separate P&L. So DeepEalth is actually an informatics business where the focus is to bring in a lot of value to our customers in the area of, let's say, cancer screening, precision medicine, diagnostics, and also oral health care. So that's kind of what we do with DeepEalth. So you see a lot of images. So images is a big part of the multimodal AI models. I'm connecting the dots here in real time. I'm almost imagining that there's real efficiencies in there. Can you explain some of the things that's going on as AI comes in? What are some of those efficiencies? Obviously, going through all the data scans, which is images, AI is helping there, I'm assuming. Can you take a little bit of color commentary on some of the things you guys are doing that drives that value? Yeah, so I think so I'll talk about four trends that's happening and one is efficiency. The other one, what we're looking at is how can we improve patient lives? That's the biggest part of our vision and what's our vision from our CEO, Dr. Berger, or our CEO, CTO, or Sham Soka, right? Is that how can we improve lives of our patients? Now, in terms of how AI is helping us, so if we look at our population, there are four major trends what we see in the industry. One is the mammal cancer, or breast cancer, what we call, second the prostate cancer, third one is lung cancer. Now you see a lot of people smoking and now we are vaping and all of that, we see that there's a big trend in the industry where people have a lot of lung cancer. And then you have the brain degenerative diseases that's also kind of been there. So in these areas, we actually have a lot of AI assets and a lot of them were built on cloud native Google, right? So I'll take it in point in case in terms of the breast cancer. So we have a product called the Deep Health and now we call it Sage Breast. So ideally what it does it, it actually gives you a scoring of the potential likelihood that the patient might have actually management tumor or not. And today with the lack of, let's say radiologists in the industry, a general radiologist can use this AI and actually read it as good as an expert radiologist in mammal. Now what we're doing is, this is the efficiency in what I'm talking about. I'm taking a general radiologist, I'm making it as much efficient as a radiologist training in mammal, huge efficiency because otherwise very, very few people, general radiologist wants to read mammal because it's very complex in itself. So there we are. Kind of an art to it too, if the experts know what to look for in the data that a general person might not know. So the screening is just, you're creating essentially the ability to. Yeah, we are assisting a radiologist to become much more efficient like an expert radiologist. That's one part of efficiency. The second part is we're actually removing large, when radiologists read it, they're seeing a lot of blacks. There are opportunities where they might miss something. The AI comes in handy there, but it actually can help you to not miss anything which can cause harm to patients. In fact, we have a study, probably we can share a link, but there are a lot of studies on how we are improving the breast cancer screening and the efficiency of it and effectiveness of it in the industry. And probably we are the largest, by the way, in the United States to do the breast cancer screening. I love how AI pulls forward these benefits, just saving lives, identifying screening early, early detection. I mean, this is exactly what we want. A pull forward, but it's not going to go away, unlike other examples. So one important distinction between us and probably the rest of the AI players in the market is the following, right? So there are a lot of AI companies building a lot of AI algorithms, but what probably they miss the market is AI should solve end-to-end workflow problem in healthcare, not just give you algorithm, because if I go and look at the Mamo or breast cancer algorithm, you might find a dozen of them in the industry. But what RedNet has done is taken these algorithms built in-house with our experts and actually have made what we call a flagship program today called Enhanced Breast Cancer Detection. It's actually the flagship program where you can do, by the way, now in Walmart. You go to Walmart, you can go and get the screening done for $40. So, but we control end-to-end, by the way. We enroll patients. We have them come and get a scan or screen for cancer. We then use AI now to basically make sure that we're detecting the right outcomes. And then we are actually also tracking patients with the comeback. So it's an end-to-end program. It's just AI is a part of it, but ideally this is what customers are looking at, of how AI is able to drive end-to-end outcome to our customers. That also adds to a lot of value into how patients are taking advantage of these solutions and making it really cost effective, right? So that expands the reach that patients can have and get those... I think that point is huge. And I think it's a great call-out because this gets back down to the cloud side of it and the data center because, okay, great. You get the screen and it's a point solution in the workflow. At GDC, at NVIDIA's conference, we were there. One thing that became clear to the world at that point is that if you have the horsepower and then you have all this resource called cloud, you now can then bring workflows and data to that workflow end-to-end and do all kinds of things, build a custom, maybe even custom silicon, because maybe you have images that require a certain kind of SLA for compute and horsepower. So you start to see this end-to-end workflow not be a one-off, but a standard in AI. And also the other case, also I'll tell you in terms of a real practical use case of how Google, assets are actually very purely in this transformation. So if you take a RedNet or a deep health as Informatics ARM, so we have a lot of our assets in the data center, right? And moving all of these images to the cloud is an exercise that we have to do. But in terms of the business outcomes, we have to do it immediately. So what it means is that if I have to run a business operation with this concept, what I just said for the screening, we might not have an opportunity like to do more later to the cloud flush and then you do all of this, right? It is about how you connect cloud and on-prem together so that we can drive an outcome as we transform the cloud. It's not like I want to go to cloud and then transform. It's about how I can transform to go to cloud but still be on-prem. So Google has lots of assets like Google and Anthos or other assets through which I can connect with on-prem and cloud and skills still can drive the efficiencies because especially for, let's say, Mamo cancer screening, what you do is you take a current image, also you can look at the five prior images, at least in writing that's what we do. So let's say you find today go to cloud, I want to get my five priors that are on-prem. Now if the expectation is that I want to move all those images to cloud, it might be a non-practical thing but Google now provides ability using Google Kubernetes Enterprise Edition to basically connect on-prem Kubernetes and cloud Kubernetes and then drive the same outcome regardless of where the data is, which is kind of a nice- So they got the hybrid and multi-cloud and was a key point but I got to ask about DevSecOps. So SecOps security, obviously confidential, I mean healthcare, you got to maintain security. Can you talk about the DevSecOps automation piece of this? How does that work? What's Google bringing into the table there? So I think so we can add to it but I'll tell you the importance of cybersecurity, right? I think so recently in our industry we have seen few incidents because of which in fact, a lot of insurance claims were held up because of the recent incident that happened in the healthcare. So and also especially in the healthcare segment where the customer is still transforming the talent to secure infrastructure assets is probably one of the most challenging feat. That you're probably afraid of ransomware too. I mean tons of ransomware tax. Correct, correct. So now cybersecurity on cloud is an essential component of how we can succeed in cloud, right? And Kamal can add in terms of the assets what we use in Google Cloud but we are working with our ISVs to actually transform that as well. And the same thing applies here. So we want to not just take security for one cloud, we want to expand it to all the clouds that we are operating in take the best of the breed, whether it is a model from working on Azure or Google Cloud or on-prem on a data center. We want to have a same security posture across the board making sure all of our healthcare data, all PHIs is safeguarded and we are compliant with all the best products. And Google is a good job on that. Google has great products which is now expanding to all the clouds. And you've got all the announcements here. Google next, obviously gonna be more generally embedded in. All right, cool. So I want to get into the deep health, fascinated by the innovation strategy you just laid out that's really compelling. You got a cloud native environment. Okay, you mentioned Kubernetes. We'll have hand Goldberg on the cube tomorrow. You have this clinical AI operating model. Can you explain how that's working and where are you going to take it? Yeah, yeah. So I think so, maybe I'll send to the base before I go to clinical AI. So what we are doing in deep health we're building what we call an operating system for ideology. We call deep health OS. Essentially what it is is a platform or ecosystem through which we're able to drive some of the efficiency to all of our, let's say, end user base. For example, today, take an example of radiologist. Radiology uses viewers, they use packs, they use different tools to actually read a particular image. And actually these are sold by multiple vendors, by the way. Viewer is by different vendor, advanced imaging is by different vendor, Pax is by different vendor, AI is from different vendor. I mean if you name it, radiologist uses five or six different vendors to actually read a particular image to get to an outcome, clinical outcome. Now same with the technologist who actually scans you, he uses another, let's say, dozen tools to actually scan a patient. Now you have a friend desk who also is a dozen technologist to actually schedule a patient. And same thing after the patient goes home. The silos are just ridiculous. So what we're doing here is that we are saying that hey customers and users, we at Deep Health, we want to transform that particular, let's say, area. But we are saying that you don't need 20 different tools. We'll give you one back one at which you'll get one user experience to be able to drive an outcome. For our radiologist, he'll have access to all of his AI, all of his viewer, all of the advanced imaging, all of the work list under one platform called diagnostic workspace which is built on top of this Deep Health OS. Same thing for technologists who scan a patient. You want remote connectivity. You want to scan a patient effectively. You want to like say, make sure that the patient is, you know, patient engagement is there for him. It comes all under one platform called technology workspace, right? So- All cloud-based. All for the cloud-based. And you know, cloud-based, but of course we work in hybrid because of the- Because the workflows are there, the machines are on the edge. So the way we're doing it is we're actually using Anthos as a means to connect or on-prem. And I'm using GraphQL as one of the technology engine to API or on-prem, like I say, database. And I'm connecting and building a new workflows now without actually migrating stuff to cloud yet. But eventually they- And on-prem is Graph, what? The GraphQL. GraphQL, yes. It's lightweight, high-performance, works well with the cloud. Exactly. Yeah, good edge in this case here. Correct. Yeah. So, the D-PAL has essentially like say, an ecosystem on which we're building all of the workspaces to drive now radiology efficiency, clinical outcome for radiologists, efficiency for technologists, and also for front desk, contact center and all of that. So ideally this becomes a one-stop shop for entire radiology business. What's the ease of use on their side? Give me a taste of what it's like on their side. What's changed now? Sounds like the back end's nice. Congratulations, good job guys. Now, the user, they got a device, they got people's faces coming in. They're probably concerned. They want to make sure that they get detection if there's a cancer. Look, I'll give you a specific use case. So for example, let's assume that you go ahead for stomach pain. So a technologist is scanning you. And you find something which is not normal, right? So typically what happens is when you are getting scanned, you basically have to get what we call order from a physician. And it has set up protocols that you need to use to scan a patient. But now a technologist has found something which is he feels that it's not normal. So you need to now call up a radiologist to say that, hey, you want to add even more sequences to it so that I can get a better shot or better image of the patient. So typically in this scenario, let's say pre-OS, what we're building, it just will literally, they have to make a call to support, they call radiologist. Radiologist that doesn't know what is happening. He might, if they can get ahold of them. Exactly. One. Yeah, and then if they are basically on-site radiologist within the hospital, then they can come in and look at it. But that is basically lots of moving around things. Now take a look, let's look at OS, how we can solve it. Let's assume a technologist is basically find something which is, he doesn't feel something is right. He can click a phone button. I need to actually give you a first available radiologist. He can click a button. He can call radiologist. Radiologist, it'll pick up a call. Green, yellow indicators out there for him to know they are there or not. And then radiologist, when he calls the technologist, he can remotely see entire scanning process that is happening with the full view of patient and the previous scans. Yes, all of them. It's one click by the way. So instantly up to speed, he's, you get real-time response. And no recall, no repeats. Because imagine this, you have to send a patient back. I can bring him back again to scan it. So the user experience is phenomenal. Modern user experience, the patient experience. What happens to patient that, you know, he's fully taken care on the table. Not something that he goes back home and comes back again. Or go out, wait, wait, goes right back into the picture. Exactly. The cycle times involve the waste and cost too on the facility. Yes. All right, so now all the scans go into the cloud. Yes. And that's where everything happens. Great, so you got a great solution. Yeah, and the user experience is the same across the product. Why isn't everyone doing this? Yeah, the reason is because, like I said, there are- Bureaucracy? No, it's a good question. I mean, there are startups, like we feel that we are a startup. Startups are the one who have the DNA to basically go and say, this is not how you look at it. This is how you change your perspective. And we are, I think so, the trailblazers and building one stack for the entire radiology. That's why we call Deep Health as a Radiology Operating System. So- And this to your point, by the way, just for our other concentration, the anti-an workload's playing here. Because you look at it as an anti-an workload, you can innovate the entire process and make those claims and deliver them. And on top of it, it's just not a technology play. RedNet runs imaging business. Imagine now I can bring in service capability to deliver that. For example, if I would today go to, let's say, an imaging chain, which is beyond RedNet, and say, I can offer you Mamo screening, turnkey, with technology, with people, with services, with clinical discipline, they'll love it. Yeah, I mean, lock-in is not about bad for the customer because the lock-in's a success because it's so good that the switching costs, why would I switch to a worse solution? Correct. And there's more images, and that's the business there. It's a very sticky solution. Yeah, and it's a little bit turnkey in the sense that if you want to buy a certain specific piece of the product, you can buy it. Now you can upgrade your subscription to another part of the products. So we sell different offerings, but ideally, people can upgrade whenever they want at their like because it's cloud native, it becomes easier for them to do it. Because we are touching almost all part of patient interaction, the whole modern end-to-end modernization becomes really key, whether it is on the web, on the phone, everything is in a single platform, a turnkey solution. This is a great example of how you guys are adding value and how innovators can innovate with Google Cloud. Quickly, before we end the segment, talk about the Google relationship of persistence. This is the kind of transformation and change management and innovation strategy, frankly, that people are looking at. Complete rethinking of how they approach the market end-to-end, data workflows, intellectual property, their business model, advancements as well. So I mean, across the board innovation, this is all enabled by what you guys do with Google. Take a minute to explain the relationship. Yeah, so for Google and even for the other Google Cloud provider, whether it is Microsoft or AWS, we have deep partnership with all of them, working directly with their product owners. What products are coming up down the line for 2024 or 2125, how does that impact our customers and how can we bring all of that to the table as a turnkey solution to accelerate the transformation that our customers are thinking about and that's what we're doing here with the Cloud. Madhu, congratulations on a great product and operating system and solution. It's changing lives, certainly. It's a game-changer and people, they get one early detection. You know, you bring expertise down to the democratized level, walking to Walmart, okay, it's turnkey. Next thing you know, you got results. If you don't have that, you have to then come in, you got to get your health. I mean, just so much benefit to society. I would say this was a vision of for, let's say, CEO Dr. Berger, Howard Berger and also our vision of for CEO and CTO Shamsoka. I think so, they put up this vision of let's simplify things for our end users, of our, let's say, customers, of our operators and let's take care of these things for them where they can focus on improving patient life and then we are here with partners at Google to kind of innovate and bring the value to our customers. That's good business. Yep. Good job, guys. Madhu, thanks so much. Kamal, thanks for coming on. Thank you. Thank you for watching the queue here. Special presentation, I'm John Furrier. Thanks for watching.