 Live from Las Vegas, Nevada, extracting the signal from the noise. It's theCUBE covering IBM Edge 2015, brought to you by IBM. Welcome back to theCUBE here in Las Vegas. This is SiliconANGLE TV's premiere broadcast. Go out to all the big industry events, help extract the signal from the noise. We're of course here at IBM Edge talking about things like storage, power, Z-Series, middleware, lots of different technology that really sit at the basis for a lot of companies around the world. Really happy to have a couple of practitioners here with us on this segment. It's Coriol Life Sciences. We've got Scott McGill who's the CEO, and Steven Cradle who's the director of product development. Gentlemen, thank you so much for joining us here on theCUBE. Thanks for having us. All right, so Scott, obviously, Coriol Life Sciences explains that you're in the life sciences business. Can you just give us kind of a quick thumbnail of the company, how long you've been around, and what parts of the business you're in? So we're a healthcare company that utilizes patient DNA to help physicians understand what drugs will be efficacious for the patient, which ones might cause them harm. So we are a fairly recently formed spin out from a 62 year old non-profit that's been operating in the Philadelphia region in the areas of personalized medicine and biobanking and a whole host of other research projects. About three years ago, we spun out Coriol Life Sciences as a commercial endeavor, focused on taking a lot of the intellectual property that had been created through a research project that had been running for about 10 years at the Institute. And we figured out how to take genetic information and really meaningfully use that now in the clinic so that doctors understand which drugs are going to work and which ones aren't, which is an amazingly difficult problem. Great. Steven, can you give us a quick sketch of the IT organization inside Coriol Life? So absolutely, so one of the great things about being part of a startup like Coriol Life Sciences is the opportunity to build things cloud-first from scratch. I'm happy to say we have zero on-premises data center. We have a lot of focus on software. We're taking a close look at Bluemix for platform as a service. And really from step one, we have an environment that can scale up and down as we need it to with a fully integrated backup and resilience. All right, so I've been, I was part of the storage industry and watched it for many years and one of the kind of high-level bullet points that gets bandied about is we're trying to take data and turn it into information to make it actionable. I mean, your business is all about data. I mean, you can talk about genomics, talk about that, but that being said, how has technology been transforming your company over maybe the last few years? Yeah, well I think we're just now getting to the point where all of the data that's out there as it relates to risk information for medications and medication regimens can now finally be brought together in a way that's meaningful to doctors and is more holistic than I think it's ever been before. Certainly the genetic information is huge. I mean, there's an awful lot of data that's produced when we do a genome screening for an individual, but that's really only a piece of the puzzle because your genetics don't tell the full story. When you take a medication, it's likely you're taking more than one. By 65, the average person's taking eight medications and it goes up to 17 by the time you're 85. So what's really important is to start to bring in all the information from all of these other sources, not just the genetics, but how these drugs interact with each other, how they interact with the particular patient demographics. If you're over 65, some drugs are contraindicated. If you are taking drugs that cause dizziness or cognitive burden together, that can cause slip and fall accidents at a high rate. So all sorts of variables have to be taken into account and pulled together in a way that's actually meaningful and easy to use for physicians. It's a massive big data challenge beyond just the genetics and that's really where we focused our attention is to try to rationalize that process for physicians. Yeah, so maybe you can unpack a little bit for us. There's, I mean, huge amounts of data. We know that doctors would have to spend more hours in the day to try to keep up on what's going on there. How does the technology and the people work together and how does your company help there? Yeah, so on the genetics side, we have a process whereby we curate medical literature to understand gene drug associations. So take, for example, Plavix, which is the second best-selling drug in the world. So Plavix, on a population basis, it doesn't work for about a third of the people that take that drug. It's, they're either non-responders to the drug or they're ultra-rapid metabolizers and so they flush it out of their system before it can actually activate and take effect for them. But without understanding that genetic component for that individual, doctors just prescribe it as usual and a third of the people get no benefit from it. So what we've done is to curate all the medical literature as it relates to Plavix and the genes that are responsible for the metabolizing of that drug and we've created algorithms that can take raw genetic information from a patient and turn that into actionable physician advice. So we do a simple buckle swab test of the patient. It's a Q-tip rubbing on the inside of the cheek. That gets sent off to a lab. They return back to us the raw genetic information and we use this technology infrastructure and all these algorithms that we've built up to turn that into an actionable report for the physician. And that's really something that would have been, as you say, impossible for a physician to do on their own. So Stephen, I know just as an observer out there, I've heard how much technology is changing so fast. The cost to just sequence a genome, how long that took and how much we had to do is just dramatically changed over the last few years, when from only the mega-rich or certain corner cases to, we're talking real soon, we should be able to sequence everyone. How do you look at that from a product standpoint, from a technology standpoint? How does that impact what you're doing at work? Absolutely, so what we need to maintain is flexibility. As the scope and depth of the source data evolves, we need to be able to react very quickly. Perhaps one day we can tell you what your risk of thrombosis is. The next day we can add on to that and give you three or four other answers that are derived from that as we supplement with additional patient data. So from the technology standpoint, we need to be extremely modular to say, we don't know exactly what the shape of the input data will look like. Even drives us towards sort of a documented, document-oriented view of the data. But in effect, we need to be light under feet modular and that's what we've distributed to Bill. Yeah, I mean, flexibility has to be key from an organization overall. I mean, how do you manage this change? How do you plan for it? I think drug development is typically measured in years. You know, IT projects are shrinking as to how fast they can change things. How does the company keep up and keep ahead? Yeah, I think the good news is, we've got Steve at the helm here who's been able to build an infrastructure that has allowed us to be incredibly agile with changing in really any number of different facets of touch points that really inform our business. So it's not only changes in how we can sequence individuals. So as you say, the cost of a genome sequence is plummeted to the point where it's now almost daily practice. The first genome cost about $3 billion to do. Now you can do a whole genome in it for about $1,000. So, and we don't even need a whole genome, right? So we do targeted genotyping and sequencing of just the areas of high variability for drug response for tens of dollars. So that's rapidly changing and there's no standardization in that industry yet. So we've kind of had to create this, almost an SOA architecture that allows us to take in these variable ingestion points and also when we produce our product, our advisory information positions, be able to deliver that in variable ways, right? We have everything from fax machines to direct DHR integration, electronic health record integration and everything in between. So for us, you know, what Steve has done is, I mean, frankly, it's a work of art from a technology perspective. It has really allowed us to stay on top of all of this change that's occurring and be able to take in new information as it's produced. New drugs may take years and years to get to the market, but there are thousands and thousands of drugs out there that we still don't understand the genetic correlation with. So that's, you know, new information in that area is produced on a weekly basis. And so we have to stay on top of all of that information so that we're always producing the best of available knowledge for our physicians. All right, so, you know, IBM's obviously a partner of yours in helping with these technologies. Are you talking with IBM? We're working with them at all on the Watson Initiative and how that's impacting things? Yeah, we, so we've kind of danced near the Watson initiatives on a number of occasions. Where we are with Watson right now because so much of what we do is algorithmically based. Watson's not fantastically suited for doing the day-to-day processing of genetic information. That's more, it's by rote, you know, it's things that we've written good algorithms based on science that can align genetics to physician advice. Where we do see a huge impact for Watson in the future for us, and we've been talking with the Watson team for a number of months now on this, is to help us to do that initial unstructured data medical publication curation. So when a new paper hits the, you know, Nature magazine and shows that there's an association between a particular gene and an outcome for a particular drug, none of that information is easily consumable by a machine. Except a machine that can do natural language interpretation and uncommon connections like Watson can. So we're thrilled with this idea that in our future is likely a Watson implementation to speed along this essentially product development cycle for us on the gene drug associations. Yeah, I'm curious how the knowledge that you gain through what you're doing, you know, how much of that is what you do internally versus what can be shared kind of industry-wide if you're talking things like drug interactions, I would think that there's, you know, tribal knowledge that could be shared. Yeah, yeah, I mean, we're, because our legacy is really from the research, medical research area, we're firm believers that, you know, we all get better when knowledge is shared, right, all boats rise with the tide. So for us, you know, we participate we're one of the founding members of the Global Alliance for Genomics and Health, which is really focused on sharing genomic information for the purposes of research and discovery and all that, because quite frankly, there's not enough money in the world to do clinical trials on every potential gene drug association. What we really need to do is really data-driven discovery work that can only happen when we have very large data sets. And that only happens with collaboration. All right, so, you know, we've only got a couple of minutes left. You know, how is, how's the technology helping you? And you know, what, I guess, what's kind of the white space if you kind of had a wish list and said, you know, talk to the executives here at IBM or at the other big companies, you know, what would you look for to help you do your job better? Well, on a fairly technical side, I think we're seeing all the evolution headed in the right direction. Containerization is something that we're very interested in. Some of the talks this morning talked about IO and parallelism and that's really a big driver of our architecture. We're not so much bound up and reliant on increasing clock speeds of the wake of architecture. We're oriented towards doing a lot of work in parallel, seeking out a lot of common combinational alternatives for a drug regimen, for example. It's almost like the traveling salesman problem where you have almost an undoubted number of alternatives. So that's our focus. So you hit a hot button on my containerization there. So are you guys using microservices today or are you using container technology? We're definitely headed in that direction. We have a pilot microservices architecture with Rustful APIs. Last question I have is, what brings you to a show like Edge? Think you've been to some other IBM shows first time at Edge? What's your thought so far? What advice would you give to your peers as to why it's advantageous to come to a show like this? To come to a show like Edge? When you get this kind of mind-shared together and you focus on one particular area of technology, it's amazing just how much variety there is out there in the field while at the same time having commonalities between the different industries. So we can talk about our application of these technologies in healthcare, but it's extraordinarily similar to what other people are doing in the banking industry or elsewhere. So there's a nice birds of a feather feel for having the conversations over cocktail hours and certainly in the sessions that I think start to energize people around these various technologies. So it's been great. All right, Steven, I want to give you the last word. If you had any recommendations for your peers as to what you've gone through here and any things you look back and say, oh, I wish I'd known that when I started with some of these technologies or road bumps that they should look out for. I guess as a general piece of advice, I would recommend to really focus on containerization, non-repeatability, no special snowflake servers. And if your environment blows up, you should be able to rebuild it quickly and that's been a big focus and an excellent learning experience as well. All right, well guys, huge changes in your industry and throughout the whole industry. So Scott and Steven really appreciate you coming to share your study with our community here. And please stay with us. We're gonna have a bunch more interviews here, day two here at IBM Edge. I'm Stu Miniman with wikibon.com and we'll be right back after this quick break.