 We're back live here in Las Vegas for Amazon web services re-invent conference. This is Silicon Angle and Wikibonds, the CUBE, our flagship program. We go out to the events and expect a signal from the noise. It's a great session with Amazon leaders, technology scaling, large-scale cloud. Everyone is moving to a direction of almost 100% Amazon. Well, that's their rhetoric. I'm John Furrier with Dave Vellante. You know, that's their position. Our next guest is Munjeet Singh, principal of emerging technologies, dynamic infrastructure at Booz Allen. Welcome to the CUBE. Thank you. Love to have you guys on. Dave and I always like bringing the big brains onto the CUBE to drill down on the stuff. We had James Hamilton on just briefly talking about large-scale, he's a tech athlete, distinguished engineer. You guys are in the trenches looking at business problems and have to deal with the technology. What's your take on the cloud? I'll see Amazon saying, you know, a little bit over the top in my opinion, but it's just messaging. Zero data center in the future. Yeah, maybe for some businesses, small businesses, medium-sized enterprises, maybe. But come on, Goldman Sachs is going to throw away their on-prem data center. You know, what's interesting is that, so we, at Booz Allen, we have a big focus on public sector clients. And we focus on private sector clients as well. But when you think about cloud adopters, you would think that, you know, certain agencies in the government would be some of the biggest, you know, critics of migrating their workloads there. And what we're finding is that some of the incremental steps that Amazon has taken over the past few years have really eased people's worries around those concerns. It started with the integration of VPC. So now you've got a client who has mission critical data in their data center that they're obviously comfortable with. What's VPC? A virtual private cloud. Virtual private, okay. It allows them to extend their data center into the Amazon cloud. And it's pretty impressive technology. And that, I think, started to bridge the gap. And then a few months later, actually a couple years later, Amazon introduced GovCloud. And so that was primarily targeted at public sector clients. And what that did was it basically made a community cloud that was dedicated for government clients. And so they had, they could risk easy at night that their tenants sharing that cloud with them were other government clients, other federal agencies. And then everybody that's working the systems in the Amazon data center are US citizens. So that helps from a compliance perspective. And then, you know, they've done recently a program called FedRAMP. And what FedRAMP does is it allows these agencies to do what's called information assurance reciprocity. So if Health and Human Services, which they did, creates a information assurance package, which essentially says the government says this is okay to operate inside the cloud, the army can go and reuse that package. DISA can go and reuse that package with very little effort. So I think a lot of these changes, you talk about HSMs that have been introduced recently, hardware security modules that allow cryptology and different levels of encryption to be introduced, are really starting to address some of the concerns. And what we're seeing, it's funny you mentioned Goldman. I can't talk about Goldman specifically, but we have clients who are in the financial services sector. Who are the smile. Who are actually, I don't think we're engaged with Goldman, but we have clients who are in the financial services and they're actually using GovCloud because they like the level of control that's been introduced and it addresses a lot of their concerns. So I got to ask you because Dave and I have been talking about this for since last week, end of last week, really, the whole IBM, Amazon court battle around the CIA deal, I think it was like, what, almost like $200 million roughly, here it takes, oh no, that was the bid was the delta, okay, close to 600 million, 600 million, yes, huge number. That's starting, that was the ante. So IBM tried to, in essence now, I'm reading court documents from what I could read, not a lawyer, is they had a, kind of like a clause in there that said, they essentially low balled the map reduce piece. Okay, so essentially it's- Misinterpreted, I've been saying this before, it's basically bidding 101 when they go to school. Yeah, this is a play, this is right out of the play. This is where you get to change what everyone makes money, that's how it works. And they down talk the features. Yeah, yeah, yeah, so Amazon wins, the judge says you're superior in all categories and Dave has a chart he put on Twitter that says, superior in all categories except for one of which they were equally good was security. So Dave- Yeah, there was one, the IBM one, which is management approach, but security was basically adequate for both. What's your take on that deal? I mean, can you comment publicly about it? I mean, honestly, it's a big validation for Amazon because this is the world's different. Now you're seeing huge federal bids being won. You know, I think it's a step in the right direction. I would, you know, I'm not involved in that type of work or that market, but I would love to see how it was constructed and what types of justifications they used in order to pursue a full, you know, public cloud provider for all their solutions in that space. That's one of the things that fascinates me the most. And then when you guys are talking about that table that you put on Twitter, so the government goes through a evaluation criteria and they rank different periods. They rank, you know, how was the demo? How was the technical structure written? You know, you mentioned management. And so they go through and you can get a blue, which is like an outstanding grading. And then, you know, it kind of goes based off that and that's how they determine who they're going to award the contract to. I'd be interested to see if the reason that IBM scored higher on management is that they were willing to take more of a managed services approach, whereas Amazon typically does not, right? They give you the platforms, they give you the tools, and they do a fantastic job of doing that. But then you've got to figure out the right kind of operational context and bring that to the table. I think that's right. I think IBM was willing to do one-offs, you know, and specials, but across the board, it was a very interesting read. I read virtually every word of the judge's ruling and essentially what, and I wonder if you could maybe lend some insight, because you know these deals better than we do, but essentially what IBM did is they, the CIA said that IBM's proposal had more risk because they had a minimum guarantee that was much higher than Amazon's, even though the IBM's price was way lower, the minimum guarantee was higher, and there's a little clause that said, if you don't hit the minimum guarantee, then we can raise prices, hence pricing risk. So the GAO's petition was sort of invalidated by the judge because there was more price risk, obviously, so you see this stuff all the time, right? I mean, and IBM knows what they're doing, they're not playing games, but I guess the question there is what you're taking all this, I mean, IBM's basically fighting vigorously, they're not just going to let Amazon walk into their backyard. How do you see this thing shaking out between the sort of traditional vendors that Andy talked about this morning and the upstart AWS? You know, it's really interesting because we've encountered a lot of this. We see a lot of these companies, in some cases there are competitors, in some cases they are companies that we team with to help solve a client's problem, in some cases they're both, right? You know, we compete in some deals and then we team up in other deals. And we're seeing a lot of what we've just been calling in, we didn't coin the term, but you know, cloudwashing, right? People take their data centers, their managed service hosting capabilities, and they call it a cloud, but when you really start to dig into the details, it doesn't have the auto-scaling capability that Amazon has where you could rapidly take advantage of elasticity and scale up machines. It may not have the levels of economy scale, therefore the cost efficiencies that perhaps Amazon has, and so when we're talking about those minimums, that's really fascinating to me because I wonder if they needed to have that minimum guarantee because they don't have the platform or the basis that Amazon is. Well they both had minimum guarantees, both Amazon and IBM for the price that they were offering. But so your point about cloudwashing, so there's also cloud dogma, and you heard it from Andy Jassy this morning, I would say he's very pedantic about what is cloud and what is not cloud. So my question to you is from a customer standpoint, how important is that? Does the customer care about true cloud? Oh yeah, so when we're dealing with our clients, it's not uncommon where we're talking to them and they have procured one of these solutions and they're not taking advantage of the opportunities or the benefits of the cloud that they thought they would. I go back to elasticity, that is probably one of the most incredible features that cloud computing enables and certainly I think Amazon does it better than anyone, that auto-scaling group is amazing and you look at some of the stories out there that they've parked, New York Times, case studies around the Amazon websites. I mean, they're just amazing. You really need to have that level of orchestration and sophistication and automation built into the cloud platform in order to tap into that. That's not an inconsequential investment to make. When Vivek Kundra was the CIO of the federal government and basically, I don't know, mandated but he suggested a cloud-first strategy, was that a turning point in your part of the world? I think it was, yeah, I think it was. I think it really got people's attention, it really sort of accelerated the whole movement to the cloud. Cloud-first was one of the memos that came out. Another memo that came out was the 20 points on IT efficiencies in the federal market or in the, sorry, the public sector space. And then I think that that really led up to the FedRAMP program I was talking about earlier. That is the main accelerator right now for public sector adoption of cloud. It's really an incredible program. It was a little bit slow to start but they're starting to really pick it up. Amazon isn't the only one that's got FedRAMP. There's other company in cloud providers that have it out there but it's really opening up the catalog that government agencies can look to to start to migrate their systems to the cloud and do so with the assurance that it's a secure solution for them. So, Booz Allen Hamilton, talk about your sort of cloud strategy. Was that a catalyst? Were you guys on cloud before that? I mean, you guys tend us to sniff out trends before they happen, you know? And as they're happening, the trend is your friend for sure. So, talk about your cloud strategy and how that manifested itself. You know, with FedRAMP, what it really did is our clients were coming to us and asking us, can you help us make sense of this FedRAMP program? There was a lot of discussion out there. People were confused by the terms, the JAB, Provisional ATO, you know, the agency specific ATO, who had the FedRAMP certification, who didn't. And so what we found ourselves doing for our clients is doing essentially a due diligence or a fact finding and we would go out and say, this provider's claiming that they have a cloud and they're in FedRAMP but really they're just white labeling instances that they had before and it's not quite the type of cloud service that you're hoping to deploy in your enterprise. And so we ended up taking a big, kind of honest broker solution, you know, provider agnostic approach, which we do a lot and we were able to help them navigate that maze and really figure out how FedRAMP would help them. Today what we're focused on is a lot around, we're continuing the security push. That's a big focus of ours because even with hardware security modules, even with FedRAMP, even with GovCloud, our clients do want to see some more stringent security controls implemented. And so we have a secure cloud gateway approach that we use which allows us to audit every single transaction that goes in and out of a cloud instance from a non-premise data center. It's very popular, it's been really successful for us in deploying that with our clients and we also have an identity analytics and intelligence solution that we're developing that once we authenticate somebody through a cloud broker, we can assign a risk rating to them and depending on their activities inside that cloud we can start to adjust that risk rating and if we need to take automated actions, if we need to quarantine that individual because it appears that they're trying to execute attacks against some of the other tenants inside the cloud, if we need to take some more manual actions and investigate it, we have the ability through a cloud engine that we've developed to kick off those actions and really execute off of it. So make it the straight, so you're layering on top of things like AWS, other specials, if you will, and security that Amazon won't do, they're not doing, typically not doing one-offs, right? You will do that. Yeah, we bring those customized solutions, we try and implement stringent, more fortified security solutions on top of Amazon just to raise that comfort level even more and to give our clients a little bit more control and one of the areas we're really focusing on that is, it's through broker. I think that's going to really keep us busy and the industry busy over the next five to 10 years. So we'll talk a little bit more about that. Well, for our clients, they're looking at it for three main perspectives. You know, the first is, you imagine that you're running an IT shop and you're responsible for standing up servers, for application owners, business owners, and like a lot of these IT shops, they start to get backlogged. So your customer base starts to spin up their own instances inside Amazon, inside Rackspace, inside Microsoft. So now as an IT owner who's responsible for that enterprise, you're starting to lose insight and lose control. People call it rogue clouds, they call it shadow IT. It introduces an element of risk. And so what the broker is promising to do is centralize all that provisioning and also give a unified management approach so that we can on one pane of glass see our instances across all the different cloud service providers, figure out what is up, what's down, you know, if there's any vulnerabilities or anything that we need to address there. And that's kind of the short term, I think, benefits where we really think the broker's going to go and the long term is introducing a marketplace. So imagine if you are TSA and you're part of a Homeland Security broker. So TSA wants to stand up a business intelligence platform, they log into the Homeland Security marketplace and see that customer border protection has already stood up a broker intelligence platform. They can purchase that appliance, they can purchase like an appliance through the marketplace instead of investing the tools, the energy and all that, they can really start to reuse those government artifacts that have been developed. So that's where our main focus is and where we really see the benefit of the broker kicking in. So I got to ask you the question that came on for our last guest from the crowd, Lloyd Strimmeling asked on our crowdchat.net ask me anything application, crowd source content. Ask James, I'll ask you now, what your opinion of cloud brokerage companies such as Jamcracker and Gravitons, do they have a place or is it wrong thinking? So you were just talking about what you see as possibilities. So what's your take on that question? We kind of boil down the broker market right now to three categories. We see that there is the large ISVs who are all used to CA, BMC, these big software companies, they're taking their management platforms and they're trying to re-bundle them as brokers. They tend to have a higher price point but the implementation time to stand up the broker is really incredible because those tools are pretty mature. Once you integrate them together and you stand it up and customize it, you can really start to experience some of those benefits. Then we've got kind of the niche market and that's where I kind of take Jamcracker, Gravitons, and Stratius, a lot of these other tools that are out there and they're very niche tools. They focus on a few elements of the broker and they do it extremely well. So if the question is do I think there's a place, I think there's definitely a place for it. It depends on what your organization is really focused on. If you're focused on standing up a marketplace, there's some tools out there that really focus on that do a marketplace well. If you're focused on the unified management framework I was talking about before, then you know Jamcracker is fantastic for that. If you're really focused on the charge back capabilities and how you're going to try and create your own market using the broker, that may be a different product. So I think it depends. And the third category is this emerging open source market. There's a few open source brokers that are starting to gain momentum. Like who? The one that we're actually most interested in is not surprisingly Red Hat. So they have a product called CloudForms and it was pitched as a kind of a unified cloud, you know unified cloud management framework but they're really starting to bolt on some of their other what they call the Red Hat Cloud Infrastructure Solutions and create a broker that's open source and has a lot of those benefits that open source brings. So those are kind of the three categories. Do you think this is a play longer term or maybe a midterm for like a service now to get into this business? I mean with their platform, it's a perfect application with app creator. You could write a killer broker. Platform or strike the brokers. It's going to be a platform worm. So who's the platform? Call Frank Slutman. No, but thanks to technical question around, okay, I buy the cloud broker argument, even just answering off the cuff to James. It makes sense, you got spot pricing, other paradigms from other industries will come into the tech world. I mean, it won't be that, it will be pretty obvious once it happens but things like that I can imagine. So the question is, what does the platform look like? Is it simply just APIs? So that's going to be interesting because if I'm going to join a cloud broker, I'm assuming there's some quid pro quo or contractual relationship where we're passing data back and forth. Trust relationships, that's interesting. I'm not sure from my technical, people will salute the same flag. You understand? Yeah, no, I think so, the platform, API certainly play a huge role into it. The platform kind of goes back to, you know, what I was mentioning earlier, it depends on what your broker objectives are. I mean, the platforms, they do not look that different from what we've seen previously. You look at cloud forums. It's basically a company called Manage IQ that they bought and it did cloud management extremely well and now what Red Hat's starting to look at is through some of the APIs and some of the other projects that they've got upstream in their open source communities bolting that on to add those other broker functionalities, other broker tools. I think one of the other parts of your question is, what is the level of agreement that takes place between the broker consumers and the person managing the broker? One of the things we're seeing in the federal government right now is, if a system integrator is going to take that broker role, how much risk do they assume for these cloud providers? So, if a cloud provider has a data spill or has an outage, is that the system integrator's responsibility as a broker? Is it just their responsibility to inform the broker consumer and make sure that they're well positioned to be able to handle that or that they're passing that communication through? Or do they actually take a financial hit? Those are some of the questions that are being worked through right now through DISA, through FAA, through GSA and DHS. So, it's definitely awesome. Would you guys, from an M&A standpoint, would you consider acquiring a broker like that or are you sort of agnostic to the technology? You know, we haven't had those discussions. A lot of times when these trends have hit in the past, we've been very successful in integrating those products together. We tend to be very agnostic. We don't want to nail ourselves into one solution because a broker for HHS might be completely different from a broker from GSA, and we want to be able to approach that client's problem with a different toolbox each time to make sure it's the best fit for them. So, I got to ask the question because Dave brought it up. Obviously, I'm just going to say it because I think the trend is moving towards where the high-end management consultants, I call the brains, helping customers transform are going to get much more technical and have platforms. So, like, CSC bought ServiceMesh, right? So, you're seeing the moves that they're making. So, they're bolting on as part of their delivery infrastructure and services. I mean, ServiceMesh, I mean, I think VMware should have bought that. I was thinking VMware would buy ServiceMesh or EMC given the relationship and orchestration. It's somewhat rare for a systems integrator to make an acquisition like that. Isn't it or not? You know, we made an acquisition earlier where we purchased a company called STI, and the reason we did that is they were extremely, one of the areas that we focused on is rapid prototyping. They had the capabilities. You talk about this 3D printer craze, they take it to a whole new level. I mean, I went down there, we were creating a device for a medical client, and they were spitting out circuit boards, they were spitting out new form factors at will, they do some basic design and produce it. And so, we acquired them because we felt that that really fell in line with what we were trying to accomplish in the defense space, in the intelligence space, you know, I mentioned commercial. So, you know, it's not, it is becoming, I think, more common. And I wasn't completely surprised to see it, but it did, I think, raise my voice. You're seeing ad agencies, for instance, buy these social listening companies because they're vertically integrating what will be automation of potential, you know, I call the arms and legs of technology. You know, you're getting it in there. So, it's the parameter there, just to, you know, speed, control, tighter integration, just strategic, all the above. For the broker? Yeah, no, for a systems integrator to make an acquisition like that, which, again, it's not every day that you do that, but you're right, it does happen. Is, I mean, or you could be the chef and just choose the best ingredients, but you choose not to in that instance, from a systems integrator standpoint, why do you choose to vertically integrate? Is it for control? Is it for speed? Is it for? It really is. I mean, that, so when we, so okay, I'll give you one example, a client with the U.S. Army, we had a choice of going between VMware or Red Hat Enterprise Virtualization, which is a KVM-based hypervisor. We decided in that particular case to go to Red Hat, because when we looked at the features that they needed, VMware, quite frankly, was just overkill for them. They didn't, they had some mission critical applications, but nothing supporting, you know, the war fighters directly. And so they didn't need some of the fault tolerance that was built into VMware. Red Hat's caught up on that now. They've done a lot of that. And VMware is still a solution that we deploy 80% of the time. But, you know, we did not want to have to come to that engagement with just a pure VMware approach. We wanted to have the flexibility to do a market landscape assessment. And it was, it was, you know, an odd choice at the time because KVM was not very popular, but it's really worked out for that client. They saved $5.1 million that year, and those savings are starting to incur every fiscal year since, so. Okay, we got a break here. We're a little bit over our time. We're getting yelled at by the timekeeper here. Great conversation. That's really important. Again, I think that, again, this is just top of the first inning I was telling Dave earlier, the whole cloud prior to this year, in my opinion, was spring training. You're going to start to see real money at the table. You see people looking at a direction, real path, although it might be bumpy right now, but that road be paving. And, you know, it's guys like you guys doing all the heavy lifting in terms of the brain power execution and helping customers get there. So congratulations. I think it's a good strategy. And great answer on the cloud brokering question. I think that's solid and answers our crowd check question. This is live exclusive coverage. The Cube here at Reinvent. We'll be right back after this short break.