 Okay, welcome back everyone to theCUBE's coverage here in Las Vegas for AWS Remars. Remars stands for Machine Learning Automation Robotics and Space, part of the reshows, Reinforced Security and the big show, Reinvent at the end of the year is the marquee event, of course theCUBE's at all three and more coverage here. We've got a great guest here, Ryan Rees, Practice Lead, Data Analytics, Machine Learning at Mission Cloud. Ryan, thanks for joining me. Absolutely. So we were talking before we came on camera about Mission Cloud. It's not a mission as in a space mission. That's just the name of the company to help people with their mission to move to the cloud. And we're a space show, you can make that, it's almost like plausible. I can see a Mission Cloud coming someday. Yeah, absolutely. You got the name. We got it, we're ready. You guys help customers get to the cloud. So you're working with all the technology on AWS stack and people who are either lifting and shifting or cloud native born in the cloud, right? Absolutely, yeah. I mean, we often see some companies talk about lift and shift, but we try to get them past that because often a lift and shift means, like say you're on Oracle, you're bringing your Oracle licensing, but a lot of companies want to innovate and migrate more than they want to lift and shift. So that's really what we're seeing in the market. You see more migration, less lift and shift. Yeah, exactly. Because they're trying to get out of an Oracle license, right? They're seeing what that's super expensive and you can get a much cheaper product on AWS. What's the cutting up areas right now that you're seeing with cloud and Amazon? Because Amazon is at their birthday, DynamoDB to celebrate their 10th birthday. Where are they in your mind relative to the enterprise in terms of the services and where this goes next in terms of the on-premise? You have the hybrid model, everyone sees that, but like you got Outpost, not doing so as good as say EKS or other cool serverless stuff. Yeah, I mean, that's a great question. One of the things that you see from AWS is really innovation, right? They're out there, they have over 400 microservices. So they're looking at all the different areas you have on the cloud and that people are trying to use and they're creating these microservices that you string together, you architect them all up so that you can create what you're looking for. One of the big things we're seeing, right, is with SageMaker a lot of people are coming in looking for ML projects, trying to use all the hype that you see around that, doing prediction, NLP and computer vision are super hot right now. We've helped a lot of companies start to build out these NLP models where they're doing all kinds of stuff. You use them in gene research, we're trying to do improvements in drugs and therapeutics. It's really awesome. And then we do some e-commerce stuff where people are just looking at, how do I figure out what are similar things on similar websites, right, for search companies. So that's awesome. Take me through the profile of your customer. You have a mix of business. Can you break down the target of a small, medium-sized enterprise, large, all above? Yeah, so mission started working with a lot of startups and SMBs and then as we've grown and become a much larger company that has all the different focus areas, we started to get into enterprise as well and help a lot of pretty well-known enterprises out there that are not able to find the staff that they need and really want to get into the cloud. I wanted to dig into the staffing issues and also to the digital transformation journey. Okay, IT, okay, we all kind of know what's turning into more dashboards, more automation, DevOps, cloud-native applications, all good. And I can see that journey path. Now the reality is, how do you get people who are going to be capable of doing the ML, doing the DevOps, DevSecOps, but what about cybersecurity? I mean, there's a ton of range of issues that you got to be competent on to kind of survive in this multi-disciplined world. The old days of I'm the top of rack switch guy is over. Absolutely, yeah, it's a really good question. It's really hard and that's why AWS has built out that partner ecosystem is because they know companies can't hire enough people to do that. If you look at just a migration into a data lake, on-prem often you have one guy doing it, but if you want to go to the cloud, it's like you said, you need a security guy, you need to have a data architect, you need to have a cloud architect, you need to have a data engineer. So in the old days maybe you needed one guy, now you have to have five and so that's really why partners are valuable to customers is we're able to come in, bring those resources, get everything done quickly and then turn it over. Again, before we came on camera here live, you guys have a service-led business, but the rise of MSPs, managed service providers, is huge, we're seeing it everywhere, mainly because the cloud actually enables that. You're seeing it for things like Kubernetes, serverless, certain microservices have certain domain expertise and people are making a living providing great managed services. You guys have managed services, what's that phenomenon? Do you agree with it and why did that come about and does it keep going? Is it a trend or is it a one-trick pony? I think it's a trend, I mean what you have it's the same skills gap, right? Is companies no longer want that single point of failure? We have a pool model with our managed services where your team's working with a group of people and so we have that knowledge and it's spread out and so if you're coming in and you need help with Kubernetes, we got a Kubernetes guy in that pool to help you, right? If you need data, we got a data guy and so it just makes it a lot easier where hey, I can pay the same as one guy and get a whole team of like 12 people that can be interchangeable onto my project so I think you're going to see managed services continue to rise and companies just working in that space. Do you see a new skill set coming that's kind of got visibility right now but not full visibility that's going to be needed? I ask this because the environment's changing for the better, obviously, but you're seeing companies that are highly valued like Databricks, Snowflake, they're getting killed on valuation so they're going to have a hard time retaining talent, my opinion, probably true, but if you're Databricks, you can't raise that $45 billion valuation, try to hire, seeing people. They're going to be under water from day one so there's going to be a real slowdown in these unicorns, these mega unicorns, decadcorns, whatever they're called because they got to refactor the company's stock equity package to attract people so they're going to put them on a flat foot and the next question is do they actually have the juice, the goods, to go to the new market? That's another question. So what's your take on, you're in the trenches, you're in the front lines. That's a great question. I mean, and it's hard for me to think about whether they have the juice. I think Snowflake and Databricks have been great for the market. They've come in, they've innovated, Snowflake was cloud native first so they were built for the cloud and what that's done is push all the hyperscalers to improve their products, right? AWS has gone through and drastically over the last three years, improved Redshift. Like I mean, it's night and day from three years ago to today. And you think Snowflake put that pressure on them? Snowflake absolutely put that pressure on them. I don't know whether they would have gotten to that same level if Snowflake wasn't out there stealing market share but now when you look at it, Redshift is much cheaper than Snowflake so how long are people going to pay that tax to have Snowflake versus switching over to a new service? That's a nice state of clean room. That's some nice lock-in features only on Snowflake. The question is, will that last? I mean, I see you smiling, go ahead. Clean room's a concept that was actually made by Google. I know Snowflake's trying to capture it as their own but Google's the one that actually launched the clean room concept because of marketing and all of that. Google also launched a semantic layer which Snowflake's trying to copy that. What does that mean to you? When you hear the word semantic layer, what does that mean? I mean, semantic layer just is really all about meta tags, right? How am I going through to figure out what data do I actually have in my data lake so that I can pull it for whatever I'm trying to do, whether it's dashboarding or whether it's machine learning. You're just trying to organize your data better. Ryan, you should be a Q-Pose. You're like a masterclass here in IT and cloud native. I got to ask you, since you're here, since we're having the masterclass being put on clinic here, a lot of clients are confused between how to handle the control plane and the data plane because machine learning right now is at an all-time high. You're seeing deep racer, you're seeing robotic space, all driving by machine learning. Swami said it today. The companion coder, the code whisperer, that's only going to get stronger. So machine learning needs data and feeds on data. So everyone right now is trying to put data in silos, okay, because they think, oh, compliance. You got to create a data plane and a control plane that makes it highly available so that can be shared. Right. Now, a lot of people are trying to own the data plane and some are trying to own the control plane, or both. What's your view on that? Because I see customers say, look, I want to own my own data because I can control it. Control plane, I can maybe do other things. And some are saying, I don't know what to do and they're getting forced to take both a control plane and a data plane from a vendor. Right. What's your reaction to that? So it's pretty interesting. I actually was presenting at a tech target conference this week on exactly this concept, right? Where we're seeing more and more words out there, right? It was data warehouse and it was data lake and it's lake house and it's a data mesh and it's a data fabric. And some of the concepts you're talking about really come into that data mesh, data fabric space. And what you're seeing is data is going to become a product, right, where you're going to be buying a product and the silos, yes, silos exist, but what companies have to start doing is, and this is the whole data mesh concept, is hey, yes, you finance department, you can own your silo, but now you have to have an output product that's a data product, that every other part of your company can subscribe to that data product and use it in their algorithms or their dashboards so that they can get that 360 degree view of the customer. So it's really key that you work within your business. Some businesses are going to have that silo where the data mesh works great, others are going to go. And what do you think about that because I mean my thesis would be, hey, more data, better machine learning, right? Is that the concept? So that's a misconception. Okay, so what's the rationale to share the data like that data mesh? So having more of the right data improves, just having more data in general doesn't improve, right? And often the problem is in the silos you're getting to is you don't have all the data you want, right? I was doing a big project about shipping and there's PII data when you talk about shipping, right, person's addresses, that's owned by one department and you can't get there, right, but how am I supposed to estimate the cost of shipping if I can't get data from where a person lives, right? It's just not good. All right, so another wrinkle in the equation is latency. Okay, the right data at the right time is another factor. Is that factored into data mesh versus these other approaches? Because I mean, people are streaming data, I get that, we've seen a lot of that, but talking about getting data fast enough before the decisions are made, is that an issue or is this just BS? I'm going with BS. Okay. So people talk about real time. Real time's great if you need it, but it's really expensive to do. Most people don't need real time, right? They're really looking for, I need an hourly dashboard or I need a daily dashboard. And so pushing into real time's just going to be an added expense that you don't really need. But cyber maybe, is that not, I mean maybe need a real time. Well, cyber security, yeah. I mean, there's definitely certain applications that you need real time. But don't overinvest in fantasy if you don't need an hour's fine, right? Right, yeah, if you're a business and you're looking at your financials, do you need your financials every second? Is that going to do anything for you? Yeah, yeah, and so this comes back down to data architecture. So the next question I asked was about a great country with the Fiddler AICC earlier, and he was at Facebook and then Pinterest, he was a data architect and built everything he said themselves. We were talking about all the stuff that's available now are all the platforms and tools available to essentially build the next Facebook if someone wanted to, from scratch. I mean, hypothetically, thought exercise. So the ability to actually ramp up and code a complete throwaway and rebuild from the ground up is possible. Absolutely. And so the question is, okay, how do you do it? How long would it take? I mean, in an ideal scenario, make some assumptions here. You got the budget, you got the people. How long to completely roll out a brand new platform? Now it's funny you asked that because about a year ago I was asked that exact same question by a customer that was in the religious space that basically wanted to build a combination of Facebook, Netflix, and Amazon all together for religious space, for religious goods and church sermons. We estimated for him about a year and about $9 million to do it. I mean, that's an A round these days, series A. So it's possible. So enterprises, what's holding them back? Just dogma, process, old school, legacy? Or are people taking the bold move to take more aggressive, swiping out old stuff and just completely rebuilding? Or is it a talent issue? What's the enterprise current mode of reset? You know, I think it really depends on the enterprise and their aversion to risk. You know, some enterprises and companies are really out there wanting to innovate. You know, I mean, there's companies, you know, an air conditioning company that we worked with that's totally, you know, Nest was eating all their business. So they came in and created a whole IoT division. You know, to chase that business that Nest stole from them. So I think often a company's not necessarily going to innovate until somebody comes in and starts stealing their lunch. You know, Ryan, Andy Jassy talked about this two re-invents ago and then Adam Sileski said the same thing this year on a different vector but kind of building on what Andy Jassy said. And it's like, you could actually take new territory down faster. You don't have to kill the, I'm paraphrasing. You don't have to kill the old to bring in the new. You can actually move on new ideas with a clean sheet of paper if you have that builder mindset. And I think that to me is where I'm seeing and I'd love to get your reaction because if you see an opportunity to take advantage and take territory and you have the right budget, time and people, you can get it. It's gettable. So a lot of people have this fear of, that's not our core competency and they're the frog in boiling water. You know, my answer to that is I think part of it's VCs, right? VCs have come in and they see the value of a company often by how many people you hire, right? Hire more people and the value is going to go up but often as a startup, you can't hire good people. So I'm like, well, why are you going to go hire a bunch of random people? You should go to a firm like ours that knows AWS and can build it quickly for you because then you're going to get to the market faster versus just trying to hire a bunch of people. Well, Ryan, I really appreciate you coming on. I'd love to have you back on theCUBE again sometime. Your expertise and your insights are awesome. Give a commercial for the company, what you guys are doing, who you're looking for, what you want to do, hiring or whatever your goals are, take a minute to explain what you guys are doing and give a quick plug. Awesome. Yeah, so Mission Cloud, you know, we're a premier AWS consulting firm. You know, if you're looking to go to AWS or you're in AWS and you need help and support, we have a full team. We do everything, resell, MSP, professional services. We can get you into the cloud, optimize you, make everything run as fast as possible. I also have a full machine learning team since we're here at Remars. We can build you models, we can get them into production, can make sure everything's smooth. The company is hiring. We're looking to double in size this year. So, you know, look me up on LinkedIn, wherever. Happy to take notes. You mentioned theCUBE, you get a 20% discount. He's like, no, I don't approve that. Thanks for coming on theCUBE, really appreciate it. Again, machine learning, Swami said on stage, you can just be a full-time job just tracking just the open source projects. Never mind all the different tools and like platforms. So I think you're going to have a good tailwind for your business. Thanks for coming on theCUBE, appreciate it. Brian Rees here on theCUBE. I'm John Furrier with more live coverage here at Remars 2022 after this short break. Stay with us.