 Hey, welcome back everybody. Jeff Frick here with theCUBE. We are live in Las Vegas at AWS re-invent. 32,000 people, I think the cloud thing is real. It's packed wall-to-wall. You can't hardly move around. We're excited to be here, our fourth year coming to the show. And we're excited by our next guest, Praveen Ragnath, the Director Cloud Marketing Explunk. Welcome, Praveen. Thank you. And with him, Gary Stokes, Director of Product Engineering, Family Search. Welcome. Thank you. So I think most of our audience knows a little bit about Splunk, but we don't know so much about Family Search. So Gary, give us kind of the quick, what's Family Search, what are you guys up to? Family Search is a genealogical organization, the largest in the world. We have about two billion records of historical records that are out there that we publish online. That represents five and a half billion people. With those records, we have 17 million registered users. We put about two million new images out a day of historical records. So we have a lot of new things coming online there to help people as they look for their families and try and learn about ancestry. That's a lot of big numbers. You're saying I'm faster than writing them down. So how long has the company been around? Well, it's been an organization ever since the late 1800s. We've had a website ever since the middle of the 1990s. Wow, so you collect data from all different disparate data sources, I assume. That's right. And then get it all digitized and make it searchable and for people to be able to go in and manage it. Yes. Excellent. So you're here at AWS. I presume you've made the move to the cloud. We have. About five years ago, we decided that we needed to figure out how to handle our hardware, our machines, how to accelerate our development. And as we went through and looked at that, the cloud became the obvious place for us to move to a platform that we could adapt as fast as our developers needed. So we've gone pretty much all to the cloud. So was it mainly just for scale? Was it for growth? Was it for flexibility? What were kind of the primary drivers that were breaking down in your own internal systems? I think it was all three of those. Particularly scale, where we vary quite a bit over the course of a week and so we didn't want to have the maximum number of machines all the time. And our development efforts were varied quite a bit and so we didn't want to be purchasing machines that wouldn't work for other development just for a specific project. Okay. And Praveen, was Splunk involved with Family Search then? Or are we coming to the story a little bit later? Oh, we were not. They've been a long time customer of Splunk. Oh, okay, great. Yeah, they actually have a tremendously interesting use case with us. What they've done is actually emblematic of digital transformation. They've had a model and they've completely transformed how they service their customers, how they deliver features to their customers. They were releasing code once a month on their website before and now they've transitioned to 900 times per day. And they've done that through AWS and they've done that through the visibility of Splunk. 900 times a day. Okay, so there's DevOps defined right there, right? 900 times a day. So that is a massive change. Not only in what the physical function, but the culture had to be a huge change to move to that. Tell us a little bit of the story of who decided to make this change. And then the part I always think is interesting is how do you sell it internally? Because we went to the developers and asked them, what are the things that frustrate you the most as you're trying to do development and get products out? And there were a couple of things that really bothered them. One was the amount of time it took from when they coded a feature until that feature actually showed up and they could get feedback on it. Which was how long, kind of? Well, we were doing, we were at once a month. Once a month. When we did it, yeah. And then the other thing that they were frustrated with was when there was an outage. It took them forever to figure out in this system of multiple services, which services that's causing that. And so we created this mindset of DevOps that said as soon as a developer makes a change, checks it in, we'll build it and we'll put it on the production site so it immediately exposes what's there. And that's where Splunk came into play because we needed something that would tell us, allow us to be checking that system and say, is it still valid all the time? That's good to say. I presume some of the pushback was, wait, wait, wait. We're going to break something, the old Zuckerberg, right? Develop fast, break fast, fix fast. So this is where Splunk comes in. This is really what you put systems in place to make sure that you could actually execute on this vision. That's right. The developers, you hit it right on. The developers were really nervous about, well, we don't want to push out code if it hasn't been completely tested. And so part of changing the culture was putting automated tests around everything. So every check-in runs through all the automated tests and then putting Splunk doing checks on the production system. So if it made it all the way through that, then we would still catch it with one of the checks later on. So Praveen, we've been going to Splunk.com for years and years and years. It's not really something I guess I think of as top of mind is that an application for Splunk. Is this new? Is it a different twist? Is it a new application? How does this fit within kind of your product portfolio and what you guys are delivering to customers? Yeah, you know, it fits quite squarely into what we do. So we have customers, as you know, using us for security use cases, IT operations use cases. That's the one we think about all the times, the security and logs and all that stuff. Exactly, and even in the AWS world, it's the same thing. It's the visibility into what's going on AWS to IT ops. Now, the key thing of what organizations use in the cloud now for is digital transformation, changing entirely how they run their business, how they go to market. Now, what's exciting to us about working with Gary is you can't deploy 900 times a day unless you have real time visibility into the impact of each and every single one of those 900 changes. So you can know what's working, what's not working. So we have customers using us in that application delivery DevOps use case. And I think this quite frankly is one of the pinnacle examples of an organization that's been able to make that transition by leveraging the visibility that we provide. So Gary, how long did it kind of take once you decided to make the move to kind of get through what you would define as transition? We made the move and we announced the development organization. Then it took us about eight months to tool up to build the tool set that we needed, get them ready. And then we started moving products over time. It's been kind of a transition because to move them to AWS, we didn't just want to pick them up and drop them on AWS. There was some rewriting. And so we've been about three and a half, four years in the process of moving things. And was the move to AWS kind of the catalyst to go ahead and let's make this other move to more of a DevOps centric way of doing things because we're going to rearchitect some of these applications? It was the move to DevOps. Oh, it was the DevOps that caused us to want to move fast enough that we were looking for a house of solution there and AWS was the right answer. All right, that's great. So law of unintended consequences, right? Always bears its head. Sometimes positive, sometimes negative. We heard in the keynote today from the guy from Italy and from Andy about some of these huge benefits that come that you don't necessarily plan for. Well, I wonder if you could share some stories of things that now that you've moved into this 900 times a day release cycle. But what are some of the things that you didn't think that now you guys are doing? That it's like, wow, really rethinking the way that you deliver service to your clients. Yeah, so two of the things we were trying to solve was how much time does it take us to get something out there? And the second one was when something breaks, how quickly do we know about it and how quickly can we identify and fix it? We used to be, if we had an outage, that that would be hours for us to figure out which services that's misbehaving and to fix that. Now we're down to minutes on that because we can use Splunksy across the system see how the correlation of the different services is happening. Right. So that's really important to us and the ability to deliver really fast has made it so that we're pretty sure we know what check-in happened if it was a change related. But culturally that's changed because we only get maybe once every month or two a change related error now. And that's because the developers are much more disciplined. They know it's going to end up really quickly on production. It breaks fast, right, though, no. And then just, again, the growth trajectory of the company, has it impacted your growth? Has it enabled you to grow faster? Can you pull in more data size, more data on the supply side? How has it really impacted the company from a strategic point of view? It makes it so that our users always have a system available. And what we've found is since all of our users are volunteers, essentially, they don't get a benefit except what they like. Then they want the system up and when it goes down they don't come back for a while. Right. And so for us to keep the number of users there, we have to always be available when they want to use it. Okay, so looking back, if you're talking to a peer who's thinking of making this move, what advice would you give them? Is there something you might do different? There's something that worked really well. You know, really, what we talk about a lot is you know, find little wins. You know, find little wins and then move the organization along. What if you could share some tips and tricks with a little benefit of hindsight? Probably a couple. One was jump in. If you keep saying, well, once we fix this two or three things, you never get there because there's always this set of hurdles that look too big in front of you. Second one was get all the data there that you need because this is data driven. Gotta make the decisions on are we improving based on data? And so what we did was we pushed everything into Splunk and then backed off things that weren't important rather than trying to identify from the beginning everything that was important. Yeah, we are alive. They are making a lot of noise here at the conference hall. You know, I'd love to add one thing to what you said, Gary. Yeah, absolutely. I was going to ask you. Yeah, which is to think big, which is something I think that you did in this whole effort. When the effort to move to DevOps and move to continuous delivery came in, you know, I don't think you envisioned going to 900 deploys per day when the process first started, but they thought big and they let the process go and they discovered what the technology enabled them to do for the business. And that mentality is what I think has driven those breakthrough results for you guys. And I wonder for me, from your point of view, as kind of the supplier of some of this underlying technology, you've seen this over and over and over. How are you seeing your landscape change as people move to this DevOps mindset, really adopted? Oh gosh, we are thrilled to be in the center of this. When just here at the show at AWS, when we see the customers that are on stage at the event, Finner CIO spoke. Finner spoke with us last year here. They do all their security and compliance monitoring for AWS as well as their application monitoring as well as their cost management using Splunk. We have family search talking about their cost management use case. Travis Perkins was a $6 billion home improvements manufacturer, well retailer out of the UK doing their operational security management for AWS using Splunk. We have Eastman Chemicals, a $10 billion plastics chemicals manufacturer doing their operations, applications monitoring using Splunk. We have large global organizations doing their DevOps application delivery using our tool to just be successful in AWS. And we're just thrilled about that. Right. Well unfortunately we have to leave it there. We are on a super tight schedule here at the show. So Praveen, thanks for stopping by Gary. Great to catch up. Thanks for sharing your tips. Basically jump in and think big. That's the takeaway. I'm Jeff Rick. You're watching theCUBE from AWS Reinvent. Thanks for watching.