>> Hi, everyone. I'm really excited to be here today. My name is Jimmy McDermott. (bright music) Excited to be talking about logging analytics and how much ChaosSearch has helped us scale our data lake. So just by way of background for Transeo, our overarching mission is to eliminate the pencil and paper gaps in educational systems. And what that looks like in reality is storing a lot of data for school districts, because everything that's on paper right now can be converted to some kind of electronic digital process. Now we're part of a new ed tech product category that's been emerging over the last few years called Readiness Solutions. We pulled together all of these disparate data points that schools are housing on students and show it to students in a really consumable and digestible way for them to understand how close am I to graduation? What am I falling off track by picking a particular class or what have you? And so by doing that, you can just kind of start to grasp the sheer amount of data that we're pulling in per student, per district, across the country at scale. and why logging started to become really, really critical for us. When it comes to just the logs themselves, its actually pretty simple but the infrastructure and the requirements around it are not simple. You have one big monolithic service, but we've got many different types of logging outputs so things that are coming from our database driver, things that are coming directly from our application layer, our networking layer and all of those are coming in to currently kind of a central repository. We offer retention for data and for logs up to our longest customers' requirement. So our longest customer's data requirement right now is holding onto data seven years post-graduation. Before ChaosSearch, we had kind of this mismanaged way of bringing all these different items together. It was truly a mess. Like we were really kind of at our wit's end looking for a solution that was going to actually bring all these stuff together. We did consider spinning up a self managed elk stack. It really struggles at scale with that retention and that historical data. It's fine for spinning something up to analyze, you know, really hot data that's hot for like a day. And then it needs to get flushed out of that system so that it can stay hot and stay cost-effective because standing up those stacks yourself is something that was just going to break the bank for us. So we were truly lost looking for the right solution. And then perhaps most importantly, in a sense that it couldn't break the bank. ChaosSearch met all of those needs and then more. We stream our logs directly from our Kubernetes infrastructure, right into our S3 buckets, which is amazing by the way, because when we were setting up our new DevOps environment, we had engineers basically saying like, "why would we do that?" Like, "why not just ship it to this?" Like, "why go to the extra effort "of setting up a Fluentd connector to move things in S3 "and they're all sold." Now, it didn't take long for them to really see the value of why we were doing that. And then the cool thing is that we don't really have to worry about those retention policies being managed by us anymore because S3 has all of that built in. Our developers can actually iterate faster now because they're able to access real life production logs around certain features, around certain capabilities that they previously couldn't. And so they can actually make decisions about new architecture components or refactoring that are backed up by data. And that's really at the core of everything we're doing. On a super tangible level, we actually some recent technical diligence that we had went way faster because we own our logs. Usually, that's not something that ad tech companies are really thinking about and so making this move actually led to a faster turnaround time for us on that tech diligence which was really exciting. For the cost savings that you get for a solution like ChaosSearch and then the fact that you layer on those enterprise type of features like Abak and SSO and these other things that are part of the platform that with a different company you would pay ridiculous amounts of money for, that's incredibly appealing for a company that is dealing with intense data security and data governance requirements, but also not a super big company, right? We can't afford enterprise contracts. So this is exactly right and it's exactly one of the reasons that we were so drawn to ChaosSearch. (bright music)