>> From Austin, Texas, it's theCUBE, covering Pure Storage Accelerate 2019, brought to you by Pure Storage. >> Welcome back to theCUBE, the leader in live tech coverage. Lisa Martin with Dave Vellante. We are in Austin, Texas for Pure Accelerate '19. And we're excited to be talking with another one of Pure's happy successful customers. We've got Scott Pedram, the storage architect from One Gas. Scott, welcome to theCUBE. >> Thank you for having me. >> So One Gas. Give our audience a little bit of an overview of what One Gas is, what regions you serve, and then dig into your role as a storage architect. >> Of course. So One Gas, we're a natural gas utility company. So we're the downstream, the inline. So we actually deliver the natural gas to our customers, residential and commercial. We operate across Texas, Oklahoma, and Kansas, and various regions including Austin. In my role as storage architect, I help, I mean, basically a one-man show. So design the storage, implement the storage, run the storage. And I also help out in other areas such as the servers, the DBAs, networking, kind of a little bit of everything. >> So you've been a Pure customer for about three years. We were talking before we went live. Give us an overview of your storage infrastructure, your IT environment three years ago, and what the impetus was to evaluate Pure. >> Sure. So we were previously an IBM storage shop. I had IBM SAN volume controller backed by DS 8000, FlashSystem 820s, Storwize V7000s, so different tiers of storage all being managed by VSPC. As is common, the warranty runs out on the DS 8000. So it's time to look at a forklift upgrade or whatever the case may be. I had a plan all in place to replace it with IBM, but we are a fully regulated utility company. So I did my due diligence and brought in some competitors. EMT and Pure Storage. Heard Pure's story, especially the Evergreen storage model, and the five and six year total cost of ownership was actually pretty close, but once you went beyond that, there was no contest. Pure won hands down. And again, as a utility company, we like predictable, flat costs. So the fact that we could do that and not have to have this multi-million dollar expense again in just another three or four years. >> So I got to ask you, so TCO, done a lot of TCO studies, and the biggest component of total cost of ownership is labor, humans. So presumably, you did a full TCO, you looked at it. I'm surprised to hear you say that the five-year TCO was about comparable because Pure is, the Kool-Aid injection says it's simpler. It's more modern. Wouldn't that save head count or at least FTE? >> It could if we were a more complex environment, but as it stood, there's me and one other guy kind of as my backup. So, you still have to have somebody to run it, right? >> So that's what I asked so sometimes CFO's will go, Wait a minute. If we're not going to reduce head count, I'm not going to accept that as part of the cost reduction. Is that what's going on here? Because we're going to shift labor to more high value activity so, oftentimes the CFO doesn't count that in his or her business case. Was that the case or did you find that because you're so small it really didn't matter in terms of the management complexity? I'm interested in your thoughts on that. >> We didn't background management complexity when we were calculating TCO. It was purely the cost to acquire the storage and then the maintenance. >> Oh, so there was no management cost? No human capital, okay. >> No. >> And so it's you and somebody else. >> Scott: Correct. >> Have you now spent less time managing the Pure than you did previously with the IBM? >> Oh, for sure. >> Okay. >> And when I first got it I was afraid, am I going to work myself out of a job? >> The Pure? >> 'Cause it was so easy. >> Okay, so, you had two FTE's managing storage. >> Scott: Yeah. >> What percent of your time, prior to Pure, did you spend managing storage versus doing other stuff? (Scott sighs) I mean a rough ballpark. >> Yeah, rough ballpark. >> Dave: Was it 50/50? >> I would say, I was maybe doing 60 to 70% doing just Pure storage before. And now it's 20? >> So you've gone from 60 to 70, let's call it 65% of your time was spent managing storage tuning, troubleshooting, provisioning LANs, provisioning more capacity, planning, all those things that, we love it. Down to 20%. >> Probably. >> Roughly. I'm not going to hold you to it, but. Well I guess we're live TV, so I will hold you to it. (Scott laughing) But that's a significant savings. You can calculate that over five years, right? Take your fully loaded costs and boom, that adds up. What have you done with that time? What are you now doing? I presume you're not just hanging out. >> No, my boss is watching. >> Publicly traded, regulated utility, somebody's watching right? >> No, of course not. No I've been able to be a lot more proactive. So helping out, like I said, with the server teams, the inward teams. Consulting them on looking further. What is our longterm goal or strategy? What's the five year plan, type of thing. Instead of just fighting fires all day. Or, you know, next week we have to deal with this performance issue that's going to be coming up. >> Dave: So you've been able to be more strategic. >> For sure. >> And one more question on this whole, there's intangibles there that everybody always overlooks, but actually when you live them they make a big difference. Has there been a quality effect? In other words, instead of putting out fires you're doing thing that are more strategic. Do you feel like you have better quality infrastructure? And does that affect your business? >> I would say better quality in the fact that it's more consistent. So we ended up sweeping the entire floor with all Pure Storage. So all of production and non-production, in our case, is all on Pure. So the consistency of the latency and the response times and the performance that you get out of the storage. There is no more performance problems. It doesn't exist. >> And in terms of workloads, I know you're running Splunk on FlashArray. Give us some picture of that infrastructure, the workloads that you're running on it. And the stakeholders I can imagine them in different departments and different functions within One Gas that are using this system and not even realizing it because it's just available, it's there. >> Before Splunk, real quick, we had one application, we went to Flash. They thought their processing was broken because it completed so quickly. (Lisa laughing) >> That's a good thought to have. >> Yeah. So they finished so fast they came back to us, it's broken, I'm like, no it's not. (he laughs) >> What's your use case with Splunk? >> With Splunk it started out as cybersecurity and that's kind of what brought it in, but it has since expanded to monitoring, analytics. We actually use it when we roll out our trucks to the field to ensure that we're meeting the SLAs. There's so many different areas where we use Splunk, I'd have to refer to my notes. >> So infrastructure ops has become this big thing, right? And automation and things of that nature? Or not quite there? >> Not so much automation yet. But we do have a plan, a project to start doing more automation. >> And other analytics, I presume? I mean, they're all about analytics, right? >> A lot of our application teams, like our web development team, they use Splunk a lot for their application monitoring and trying to be proactive on that. >> Thinking about the security use case. Security practitioners often tell us, well, we get inundated with incidents. We don't have the time to sort through them all. Does having Splunk on an all FlashArray, high performance all FlashArray, does it affect the response of the security team? Or how does it affect the business, the security side of the business? >> I'm not able to answer that directly, but I can say that I have seen them do a lot of select all type queries, where they're just searching for a needle in a haystack, type of thing. And previously when we had multi-tiered storage those queries took forever, but now that it's all Flash, it's really quick. >> So they spent more time waiting than they do now. I mean that could be a two edge sword. Maybe they more stuff to sift through now. (he laughs) That's somebody else's problem. >> Well the data security is critical because your dealing with customers' data, right? And almost every month we hear about data breaches in the public. Whether it's a bank, or it's a social media platform. Unfortunately they're becoming quite common. But when you're dealing with personal customer data that's a big concern. Some of the things we're hearing Pure talk about is what they're doing with data protection and data security. And also kind of this sift from not looking at data protection as an insurance policy as much as it's an asset because you have so much information, you're storing it for longer, more and more customers, more data. How is that that being reflected up the chain, even up your chain of command and to the executive folks in terms of being confident that what they have your customers data running on in those three states that you talked about, is on a very solid secure platform? >> Well, security, it requires multiple layers. So Pure having always-on encryption is a big help. So if we do have, you know, a failed module that has to be replaced. I don't have to worry about making sure that it's securely erased, destroyed, and all that. 'Cause without the encryption key it's virtually crypto erased. And then of course we have all the security agents on the servers and the applications and our security cyber team managers, all of that. >> And what about cloud? What do you do in cloud? What's the strategy? >> We do cloud where it makes sense. For instance ServiceNow and O365 we're customers to both of those. >> Dave: So SaaS stuff. >> And mostly SaaS. In my opinion doing cloud is doing a lift and shift. And using cloud as infrastructure as a service doesn't make a whole lot of sense. For us anyway. As a utility company we're very pro-capital. So if we just shift that to another provider that's all operational. >> Whereas, take ServiceNow for example and change the operational model. Right? And you had a clear business impact where it wasn't a lift an shift. It was a transformation really. >> Exactly. >> Where do you want to go with Pure and storage infrastructure? It's just like, I just want it to work. I want it to be rock solid, dirt cheap, highly available, you know, high performance, or are there things that you would like to see Pure do that can help drive your business? >> Well I think the announcement today of the FlashArray//C is what I'm probably most exited about, in that I've already asked my business partners to get me some pricing, some quotes on, can I use that for my backups as a back up target? Instead of, you know, the underlying SaaS datadisks. So that's exciting for me. The fact that it's going to be the same software that I'm used to, that's all a plus. >> How are you protecting your Flash arrays today? >> We're implementing Commvault right now So we do leverage Commvault. It's called IntelliSnap. So basically it does a Pure level snapshot and then we can mount that on our media agents. >> Okay, so, using FlashArray//C, that's the right model number, I think. So obviously you want to use Flash, if it's cost effective, for everything. If it's cheaper than spinning Disk why not use it? Do you see any advantage, in theory, for recovery speed? For sure, yeah, absolutely. I mean, if you need to do a fast recovery, I mean, it's on Flash. But with what I'm looking most forward to though is even the ingest of the data, the initial backups. If there's a lot of, you know querying and trying to figure out what's changed and what's not, that can be a lot of disk thrashing on traditional spindle drives. >> So let's look into the future a little bit before we wrap here. You've been a Pure customer for three years now. Presuming you've done some upgrades and swap outs of controllers in that time? >> Not quite yet. In the coming months we will have our first ever green controller swap. I've actually had a failed controller. So effectively the same process. Where one controller's down and didn't have any issues with performance or, >> No downtime, no disruption. >> No downtime. Absolutely not. Even upgrades where they, you know, take one controller down and upgrade it. I'll do those during business hours. >> Are you comfortable with the, go ahead, sorry. >> Just because there's no performance degradation whatsoever. >> So you're obviously comfortable with the architecture. You seem like a pretty happy customer. Some of the critics will say, it's a duel controller architecture, that doesn't bother you? >> No, not at all. (he laughs) >> I had to ask with a straight face. What would you like to see Pure do? If Charlie G. and Carl are sitting right here, what's the one thing that I could do to make your life easier, what would it be? Besides cutting price, you can't say cut price. >> Yeah. You know what, that's a great question. I think what I would have been asking for, top of mind, would have been the lower tier, what they came out with today, the C. >> You know, another criticism from some of the competitors is they don't have tiering. And when you talk to Pure about it they go, oh, we don't need tiering, we don't believe in tiering. What are your thoughts as a practitioner? Would you want to have a tiered array, like high performance Flash, lower in the same array? Or is this not something that is necessary? >> I don't think so. I go back to the consistency. You know we have all of production on Flash now and it's, I don't have to worry about performance. Whereas before I was constantly having to monitor and manage you know, is all the right stuff on the right tier, and it was a headache. >> So automated tiering wasn't so automated? Is that a fair statement? >> It worked fairly well, but there were some cases where it didn't. >> Yeah. So you're better just throwing it at Flash and it'll take care of itself. >> Yeah. >> Dave: Cool. >> So you've got a foundation now that's going to allow One Gas to evolve continually and we look forward to hearing in the next year or so when you go through that first big evergreen upgrade, how that goes. But it sounds like you've made the right choice and the foundation that you've got is pretty strong. And so many other layers of the business are benefiting and they don't even know it. Because as you said before, on of the constituents thought something was broken, it was that fast. >> Correct. >> So well done on your decision. >> Thank you. >> Thank you so much, Scott, for stopping by theCUBE and talking with Dave and me about what One Gas has been doing how you're succeeding and we look forward to hearing more of your success. >> Thank you. >> Dave: Great to have you, thanks. >> Scott: Appreciate it. >> For Dave Vellante. I'm Lisa Martin. You're watching theCUBE, from Pure Accelerate '19. (upbeat electronic music)

(upbeat music) >> Narrator: Live from Silicon Valley, it's theCUBE covering Google Cloud Next 17. >> Hey welcome back everyone. We are here live in Palo Alto for two days of coverage of Google Next 2017 special coverage brought to you by Intel. I want to thank Intel for sponsoring our editorial coverage of Google Next. It's a cloud service provider. This is a huge opportunity, cloud is changing the digital transformation and I want to thank Intel for that. Breaking down the coverage going into the realities of cloud, our next guest is Pedram Abrari, who's with Pramata, Chief Technology Officer, you guys do a customer digitization of cloud platform based in Silicon Valley. You're a veteran former entrepreneur, welcome to theCUBE coverage of Google Next. >> Thank you John. >> First tell us about what you guys do as a company. I know you guys have an interesting story because you're in the heart of the cloud game relative to operationalizing it-- >> Pedram: Yep. >> It's complicated in being an enterprise cloud solution. There's nuances there. There's some tripwires. There's some landmines, whatever you want to call that. >> Pedram: Mm-hmm. >> What do you guys do, let's do a quick background. >> What Pramata does is we are a B to B platform for large enterprises, such as NCR, HP, CenturyLink who have hundreds of, in some cases thousands of customer contracts and don't have a handle on their contracts. We digitize those contracts and those customer relationships and we layer intelligence on top to allow key decision makers in those businesses to have a single unified and up-to-date view of this data of each customer relationship at any point in time. Layering on top, building data and CRM data and MDM data. >> What's interesting why I liked that you're here is that it really hits the theme of Google Next which is data, datasets, machine learning, AI are pointing to a new model of how software's changing applications, right. >> Pedram: Mm-hmm. >> So you guys are at the middle of this digital transformation-- >> Pedram: Yeah, yeah. >> It's a whole new paradise, not like the classic, you know, linear thinking of supply chain or CRM kind of thinking. You guys are truly data driven and this teases out the complexities-- >> Pedram: Mm-hmm. >> What's your thoughts because again Google is clearly going down to the enterprise level. >> Pedram: Mm-hmm. >> As is Amazon, little bit ahead of them in terms of progress, but this is the trick everyone is doing in the digital transformation. I want to leverage my data, I want to move to a cost effective infrastructure. >> Pedram: Mm-hmm. >> Or it could be a startup saying, hey I want to get into the game and I want to innovate on a feature, and then there they are. They could be the next Snapchat out there watching. This is important, but it's also hard. What's your thoughts on the landscape of this opportunity? >> Well, cloud computing definitely changed the game for high tech startups, in a big way. When infrastructure as a service first rolled out with AWS, that's kind of the tip of the spear. The virtualization of hardware was a big game changer because as a startup to even get in the game, you had to have millions of dollars worth of investment in just hardware and software. In every two or three years, you had to renew all your hardware and software because they were out of date. So before you could even focus on your core competency, there was all these layers of investment and all the talent that you had to attract, just to deal with getting a cloud software up and running. With cloud computing, particularly with infrastructure as service, it changed that game. Virtualized hardware, and it allowed a lot more companies to have access and the ability to get into the game that couldn't previously, but the story doesn't end there, that's just the beginning of the story. Because to get a class software really up and running, you still have to have a team, traditionally it used to be IT teams, but the evolution has come now we have devops teams for good reason, who have to build a lot of initial plumbing on top of the infrastructure as a service until your cloud can be up and running in a scalable, cost effective, elastic fashion if you want. Yeah. >> Tell about the scale piece, because what is interesting is you have a lot of experience in scaling with the cloud. >> Mm-hmm. >> This is the main thing that people are leveraging with the cloud is that I can scale up pretty quickly. >> Mm-hmm. >> Scale up and scale out and then the complexity is the digitization piece, which is more specific to the enterprise. What are some of the challenges that you see with scale, because this is something that needs to be factored in on the design side. >> Mm-hmm. >> So digitize, oh yeah I want to digitize my entire company. Okay, sounds easy-- (laughter) But the scale piece is important because you now have scalable stuff. >> Right. >> How does it all work? >> Cloud software, early on in the cloud days, you know we had IT teams and we had developers who were really enterprise developers and they looked at the world with those glasses on. Very shortly thereafter, as soon as the first cloud software was up and running, people realized, what a minute, the old way of building software just doesn't work anymore, you have to rethink, this is where devops developed where it was a culture of developers and operations all working in concert, always designing software for scale in the cloud. It's a very different paradigm. Things such as transition from stateful services to stateless services to microservices, it all continued to turn services into things that could run and spun up and run across a large cluster of servers, as opposed to something that only scales vertically on a single box. If you think you have a service that you can throw on the cloud and you can magically get the benefits of that, and costs get lowered, I'm here to tell you that if you don't play your cards right, it blows up in your face very quickly. >> Give an example, cuz this is the trade-off, back to the trade-off conversation, right. >> Yeah, yeah. An example is if you have software that doesn't scale horizontally, that is not elastic, it doesn't scale and it only scales vertically, and you throw it in the cloud, and the more load gets on that software and that service, the only way to go is to keep getting bigger and bigger boxes that are available on a AWS or on Google or on Acer. The larger the box, the more expensive it becomes. The whole premise of cloud computing was commodity boxes (laughs) and things that could scale this way, and you really are basically going back to the same old problem you had on the enterprise side. Having to get bigger and bigger and bigger boxes. That can really blow up in your face in terms of the cost, that people would be shocked the kinds of bills that they can receive from some of the cloud vendors if they don't manage and contain their problem effectively. >> We're with Pedram Abrari who's the CTO of Pramata. They bring up an interesting point, I want to jump in and just kind of double down on that because the classic IT enterprise conversation in the heyday of enterprises it was developing was the sharp thin, the tip of the iceberg. What you don't see under the water, is the hidden costs, right. >> Pedram: It was massive. (laughs) >> The total cost of ownership has always been a big issue and if you look at things like OpenStack, for instance, great on paper, great philosophy, but-- >> Pedram: Mm-hmm. >> The total cost of ownership has really kind of crippled that from being, other than anything more than infrastructure as a service. So there's trade-offs for an enterprise-- >> Pedram: Mm-hmm. >> When the look at the total cost of ownership saying, I'm just going to throw in the cloud and run multicloud and everything's going to be managed perfectly and there's manageability and the security, I'm all set. No. >> Pedram: No. >> Or is that, I mean-- >> So first of all-- >> John: Why is that so important, because there's some trade-offs specifically here. >> There is ... First of all multicloud, cloud neutrality in theory sounds great, but it comes at a very expensive price. If I'm running on Google, or if I'm running on AWS and if I commit to running only on AWS or on Google or Acer, for that matter, I have the opportunity to leverage some of the managed services that are offered up by the vendor and they have the world's foremost experts at running some of theses services. Let's say your software requires a relational database, if you're going to be cloud neutral, you have to host that database, deal with backup recovery, scalability, fail overs, all of that overhead associated with that, which means you have to hire world's foremost experts at doing these things and you have to attract them, you have to pay them and on top of everything else that's associated with having to anticipate the heaviest load of your system, and always planning for that, if you can leverage the Google Cloud SQL, or if you can leverage AWS RDS-- >> Google only runs MySQL, they don't run anything else. >> That's true, that's true, but AWS does. >> Yeah. >> They have a plethora of different databases. >> It was good to go to AWS in that case. >> Well, if you're starting from ground up, and you're a startup, committing to MySQL is just fine. (laughs) >> John: Yeah. >> If you already-- >> Which is why Google's doing really well in the cloud narrative piece. >> Exactly, exactly. >> Enterprises who have other databases, other relational databases. >> And so if you're already sitting on top of a legacy that you have to support, then going to AWS might be easier. But AWS has its own complexities because it is a massive service, a massive ... Has a lot of APIs, it has a lot of complexities so you have to deal with all of that complexity. Even the billing side of AWS has a whole economy all to itself, there's all these vendors that exist just for managing AWS cost, so having a model like Google, >> John: Yeah. >> which is just a lot more simplified and kind of reduces the explosion of complexity that you potentially deal with on AWS side, may work just well for a lot of startups. >> This is really an important point I think, because this is something that's not being covered much in the press or in the analysts community is that everyone certainly talks about lock-in, oh the roach motel, you can check-in, but you can't check out. Now I've heard that-- >> Pedram: Mm-hmm. >> Been called to Amazon and everyone else, the lock-in, but if you at what you're saying is interesting. You say lock-in actually in contrast to say the opportunity of leveraging, say manageability-- >> Pedram: Mm-hmm. >> And security. >> Pedram: Mm-hmm. >> It not a big deal given the fact that you don't want to build those services. >> Pedram: Exactly. >> If you go hey, I'm fully neutral cloud where I'm going to have multiple workloads, then it's on me and IT to build the software fabric for manageability. >> That's exactly right. >> So the risk is if it's not available, (laughs) if there's no software that does that, that's the risk. >> It is the risk. As a serial entrepreneur who has done numerous startups, one of the key aspects of doing a startup is focusing on your core IP and your core differentiation. Your core IP is not how to run a cloud software, it's other peoples IP and you should leverage that. Platform as a service is a way to leverage that and you give up some control, you fall into a platform as a service, and for that matter if you want to fall into a platform as a service, you can fall into a platform as a service on AWS or you can do it with the Google app engine or you can do it on Acer, but you can basically see which one fits your needs and your profile and your software best and just give up control for productivity and for cost reductions and also you gain from all the expertise and best practices they have developed around security and audit and all the ramifications around. Basically making sure that you take care of your customer data safely and securely, you don't expose them to risk. >> This is interesting because it makes the cloud argument more about the beauty's in the eye of the beholder or whatever the enterprise thinks is best. >> Mm-hmm. >> If it's cloud native, that might be Google, but then it's an opportunity for the vendors to differentiate-- >> Mm-hmm. >> On some certain services. So I get that, but the question I want to ask you is for the folks watching who are in the enterprise trying to squint through all the complexities, hey I'm on a digital transformation, I don't know what's what, I'm seeing Google say this, Amazon says this, this is apples to oranges, what's in for me, I have my own enterprise. So that's an interesting conversation, so the question is what would you advise enterprises to evaluate when to go with Google, when to go with AWS, when to go with Oracle or IBM. There's a variety of different choices. When do you evaluate that trade-off factor of with the leverage, how do you advice that? >> It's a tough nut to crack. Before you even move to the cloud, you can still do some soul searching internally and look at the good, bad or ugly of your own software. What are strengths? What are scalability issues? Can it scale horizontally? Can it only scale vertically? With that in mind, then you go and evaluate the options that are available out there. If you're never going to leverage any of the native cloud services that are offered out by AWS or Google or Acer, and you only wanted ... Let's say you want to be completely doctorize and containerize and you really want to kind of follow that model, maybe these services don't matter to you and you're willing to take on all the other responsibility and manage all the services. So you really have to ... I would strongly advise that you gain and go to cloud experts, who have done it before time and again, and seek their incites and advice and not jump into the deep end of the pool thinking that oh, it's just cloud, anybody can do it. >> Question for you on, say Google, for instance, say that you and I were called into that Diane Greene's office and they said, "Hey Pedram and John I want you "to advise me. We really have good dev developer empathy," we talk about this in our last segment, developer empathy-- >> Pedram: Mm-hmm. >> "But we don't have a lot of empathy for enterprises. "You guys are experts in the enterprise, what should we do "to empathize with the enterprise better?" What would we advise them? What would we go in and say to her and her team? >> I would say start with the pain points of the enterprise, right. Before the enterprise can even consider moving to the cloud, their biggest and primary concern is security. They have to make sure that they can trust you and of course that has really over the years has been chipped away at, the old obstacles are following one at a time. But really being able to speak their language and get them to be comfortable that they're following best practices in their very solid and secure environment. On top of that, help them with all of their audit needs. Everybody wants to get certified. (laughs) And a lot of that when you actually move to the cloud, if you have a Google or AWS on a checkbox, a lot of those questions that auditors ask, go right out the window. So that is a helpful factor. But helping them along those lines and also cost factor. A lot of people don't know what it's going to cost. >> Yeah. >> Cost calculators and all that stuff are good and great, but they only go so far cuz there's a lot of hidden costs that you don't associate with it. A lot of it can come in the form of talented expertise. A lot of it comes in the form of just paying for services. >> John: SLA too. >> SLA, yet. >> SLA is a huge one. I would say to Diane, "Look at being a price leader, "and certainly you have great pricing, "but I don't think the enterprise is price sensitive, "I think they're SLA sensitive." >> Pedram: They are, right. >> That's kind of their weak spot, a little bit here. >> It is, and of course now Google has a little bit of advantage to bring to the table with what happened to AWS last week-- >> John: Yes. >> But again if you take the big picture of the SLAs that are offered up by any of these cloud platforms, compared to what you could do internally hosting your own services, with your own IT team, I'll bet you they'll beat your IT team every day of the week, twice on Sunday-- >> Yeah. >> In terms of SLA. So I wouldn't be afraid of moving to the cloud and again hiccups happen to everybody an anybody, but-- >> Pedram one of the things that we say clearly this year at Adibus, we've done all the live broadcasts at Adibus for years. But this year it what was clear is that the speed of which Amazon has been innovating services, and Google needs to match this cadence as well on their side for their architecture, is one of those cases where they're doing it faster than the IT guys can do it. So it's the same argument that open source is a great value because open source is moving the needle faster-- >> Pedram: Mm-hmm. >> Than homegrown teams could do on IT, so that's an opportunity to leverage that to focus on the core competency of the Internet. >> Absolutely. And then one of the other things that people overlook, when you leverage an AWS RDS service, what you gain is not just what they have at the time, what you also gain all the improvements that happen over time, on their behalf, on their side, where they keep increasing their throughput and performance and scalability. AWS just came out with the Aurora Service, which is effectively like a ... It acts like an elastic relational database, which is a concept unheard of. Imagine trying to replicate that internally. I mean it is things that the level of expertise they bring to bear, and the level of resources that they bring to bear to really solve these complex problems, far outweigh anything that we would have in our company to be able to address those same challenges. >> Pedram, great to breakdown some of these trade-offs, this is the nuances of the enterprise, being empathetic is to really understand. The buy to build kind of concept versus when do you want to leverage your core competency, when do you want to shift that-- >> Pedram: Mm-hmm. >> Capability to a cloud or certain clouds certainly the criteria. Really appreciate you taking the time. Take a minute to talk about your company. >> Pedram: Mm-hmm. >> What are you guys doing, cuz you guys are in the middle of the digitization-- (laughter) >> Physical transformation, and it's not that easy. >> No. >> What are you guys doing for customers and what's your competitive advantage? >> So what we do is, we have a lot of large enterprise customers, who typically have hundreds of thousands of customer contracts, that nobody ever looks at or reads or your only reading an army of lawyers to really comprehend and understand, and this is an obstacle to making good business decisions to grow your company. Large enterprises, much like smaller enterprises, need a up-to-date view of their customer relationships, which starts with the customer contract, which is where we come in. We digitize the customer contract and we extract key information out of it, the information, not all the legalese and noise, but really-- >> John: But the core data. >> The core data, the core key decision making data that you need to have to interface with a customer. We extract that out and make it available to you in an environment that is accessible by anybody, not just lawyers. On top of that we bring in data from across your enterprise about that customer, whether it's your billing systems, your CRM systems, or MDM systems, you name it, we can bring all of that data, layer it on top of your contract data, and on top of that, introduce additional layers of intelligence where it tells you what is the most up-to-date aspect of your customer relationship information, and that allows you to make real-time important decisions that over time your finance teams and sales ops teams can really maximize the relationship. >> This is classic data-driven, where you're taking core data about the customer and contract, they pay for stuff they haven't ... Key data in their system of record, if you will. >> Pedram: Mm-hmm. >> And kind of sharing it into other systems, sounds like it's perfectly poised for machine learning and AI, is that where-- >> That is our secret. That is our secret sauce. Trying to ingest and digitize hundreds of thousands of contracts, cannot just be done manually (laughs) clearly. >> It's not just the sales thing, but renewals, more of operationally-- >> Renewals is a big issue. There's massive operational impact, there's upsell impact, there's a lot of ... Our customers gain after adopting us, millions of dollars in lost revenue potential where they are thrilled to tell us about it, like we have found all this money we didn't know we have. It's kind of like having on top a knowledge base of data and big data, everybody knows there's information there that we could use, but to tap it, you got machine learning-- >> Cross-pollinating core data and making it addressable for other apps. >> Precisely right. >> Okay Pedram, thanks so much for coming and sharing your perspective. Breaking down the two days of special coverage of Google Next, this is theCUBE, live at Palo Alto we've got folks on the ground. Our reporters, our analysts will be calling in and of course we've got an exclusive scoop with SAP, we have one of their top executives who runs the Palo Alto entire facility, all the folks who came in from Germany. Had a chance to sit down with SAP, that's coming up shortly. Stay tuned for more coverage live from Palo Alto for Google Next 2017 in our studio. We'll be right back with more after this short break. (upbeat music)