>> Announcer: From San Francisco, it's the Cube, covering PagerDuty Summit 2019. Brought to you by PagerDuty. >> Hey, welcome back everybody, Jeff Frick here with the Cube. We're at PagerDuty Summit in downtown San Francisco. It's about a thousand people, fourth year of the show, third year of the Cube this year, happy to be back. Ironically, (laughs) a couple weeks ago we were at Sumo Logic Illuminate down the road by the airport, and we're excited to have somebody from Sumo here to talk about how do these platforms work together. So, returning again is Jonathan Rende SVP products at PagerDuty and joining us is Bruno Kurtic. He is the founding VP of product strategy for Sumo Logic. Bruno great to see you, Jonathan welcome back. >> Thanks, for having us. >> All right so Bruno we were just at your show, now you got to take a little bit easier, probably quite not as many responsibilities. We'll talk a little bit about your relationship between the two companies cause from the outside looking in, looks like there's some redundancies, it looks like two platforms, it looks like where's my single pane of glass but in fact there's a real synergistic opportunity to work together. >> Good question, so they are two platforms but it's entirely synergistic. You know between the two technologies, PagerDuty and Sumo Logics, we sort of helped our customers who run mission critical products and services that serve their customers in fact, number one, get information from their systems and applications to understand what's happening in them and then leverage our two platforms to resolve those issues, make sure those applications are running, that their customers are happy, that they're delivering the services that they are there to deliver to them. >> And I know Jon you got a long list of great companies that you guys work for and you said it's a really key part of the company strategy. >> Yeah the ecosystem that we work with with one of our favorite partners, Sumo Logic, we use Sumo, we're a big customer of Sumo Logic as well and it's really important all of the telemetry, all the machine information that's coming in. Again the part that we play as in that is how do we orchestrate people to get work done when things go south? And how do we get the right people on and give them some information about what they're doing to help triage what they're doing. >> Right. >> So the two work together really, really well. >> So what are the themes at both keynotes? Ramin's keynote as well as Jennifer's is data. And the fact that you guys both have a giant proprietary data set about machine activity and people activity from running these businesses. I was teasing on Twitter an overnight sensation ten years in the making that you can leverage to deliver more value. So, as we look forward, data's been important but, right, now all the hot topic is machine learning and artificial intelligence. How are you now taking this next gen technology and applying it to these giant datasets to offer kind of proprietary insight to your customers. I'll start with you, Bruno. >> Sure, so there's a massive amount of data, right? It's growing at a rate of Moore's Law so there's more data than any human could cope with. And so our task at Sumo's is figuring out what is that data trying to communicate to you? So we spend a lot of effort on machine learning, pattern detection, advanced analytics, to help our customers sort through that massive amount of data to understand whether their services are available, whether they're performing, whether they're secure, whether they are compliant, and we boil that up into a set of insights that we then feed downstream or upstream in this case to PagerDuty to help those people who are responsible for those services do the work to make sure they're restored and working well. >> And I guess to compliment what Bruno is saying, one of the things that we're doing is we're also ingesting a lot data, a lot of machine data from monitoring products and from service desk products, other kind of sources of data because that also informs who needs to get engaged when a system goes down? And then what do they need to do in order to fix it? And so it's all context it's all data and how we can help narrow that down. We had a really interesting statistic this was earlier this year where we were looking at per responder how is this growth of interruptions and alerts, how is that trending? And now compared to just a couple of years ago it's about three times the amount of noise that's coming at them now per responder than three years ago. So, clearly the people on the end of this are getting overwhelmed if we don't do something intelligently (laughs) to make sense of it for them. >> Right. That's interesting cause it's really a lot false positives, (stammers) I don't know if that's the right characterization but certainly too much to prioritize and an overwhelming amount of data for a person to try to filter, so you're really trying to add that intelligence on the front end so hopefully the right problems are getting surfaced and not just this broad (laughs) base of false positives, or minor positives maybe. >> Yeah, it's funny you say false positives because one of the concepts that we have is there are you know, alerts and incidents that need to be managed, but then there are un-actionable alerts and incidents. Things that really shouldn't be bothering you. So you have to walk that fine line between what do you act on that you should take action on and what are the things you shouldn't take action on and kind of ignore? And so we use machine learning to do a lot of that work and filter out the bad noise and bring the important information in. >> Yeah, I wonder if you have any thoughts, Bruno, on how much of that filtration needs to happen (laughs) to kind of quiet down this tsunami that's coming over the transom. >> Well on our terms it's, you know, every one of our customers send us billions of records per day, literally billions. >> Jeff: Billions of records per day? >> Billions of records and so figuring out what matters amongst those billions of records is a hard job. There's a lot of false positives, false negatives that need to be sorted through, before it even gets handed up to the upstream technologies like PagerDuty, right? So, we spend a lot of time doing outlier detection, doing predictive analytics, doing sort of pattern detection, machine learning type of techniques to make sure that the stuff that gets bubbled up has as few false positives and as few false negatives as possible so that the insights that intelligent actions that need to be taken are most appropriate and can be prioritized and handled by a small team of people who own those actions. >> Right, it's funny you say billions and billions. I have a digitalization challenge, I keep throwing out to people and there's yet to be, I've yet to get a great response which has shown me a billion, a billion piece dataset in a visualization that I as a person can look at and comprehend what the heck is going on. Beyond something as simple as you know, half of them on this side and half of them on this side. I mean we're not wired for that way. We're not wired to be able to take in billions of data points. It's just not, it's just not going to happen. >> Just for that context we actually, we analyze a quadrillion records a day. So talk about billions and then you know many more orders of magnitude than that, it's, those are numbers that are hard to comprehend, right? We don't think in those numbers, right? It's really hard to humans to grasp. >> So, so how do we keep up? I mean, how do we keep up? I mean it's kind of a bigger problem, but you know as much as anybody kind of exponential growth of this data. We're barely getting into IOT and industrial IOT and sensors on everything at the house and on our clothes and our shoes. You scared about keeping up? Can we keep up? What do you, you know, kind of, how do you see this crazy trajectory on the data? We have to kind of gate it somehow? >> So from my perspective there is no sense in being scared of it, right? A digital business generates data, we all got data that can't run. So the task is to capture it, analyze it, to understand it and serve up intelligence from it, right? So our task is to keep pace with that growth and build resilient scalable systems with the analytics that are required to understand it, right and so you know we can't shy away from it, so whether we like it or not. >> Here it comes (laughs). >> It's not an easy task but we can't walk away. >> Right right, and then the other just crazy increasing complexity. No, thank you. (laughing) Is on your guy's side, really is the variety. I mean we used to talk about the old big data big three you know variety, and veracity and velocity. You know the interconnectivity of all these systems is also the thing that's growing so exponentially and so when something does break the ability to find what broke amongst this huge potential is really a hard and growing problem. >> Yeah, it is and that's why it's sitting in the middle of an ecosystem of a lot of different products that will give and send off to telemetry that we have to look at. It is really important. You know, it's almost as if the information that we're always looking for on the PagerDuty platform, it has to be items that really are actionable by a person which, you know, if you look at the information that is flowing into Sumo Logic, it's even in some ways very broad. And so it's a wider funnel, we have a narrower funnel of kind of information but they're both very complimentary at each other cause one is that humans need to act on in the moments and the other one is how do I analyze in a broader sense? >> Right. >> Even a bigger range of information so both are so critical as a part of that whole ecosystem. As I was saying, we personally use Sumo Logic as a big part of how do we actually triage actual incidents? We built tons of libraries in the Sumo Logic product so we can make sense of even a broader set of information flowing in from all of our logs in some of those critical moments. So yeah, it's great synergy. >> Good, good, well I'm glad you guys are working on this big data problem cause it's a big hairy one. >> Jon: And it just keeps getting bigger. >> And the customers only benefit right? >> Yeah. >> Well Bruno, Jonathan, thanks again for taking a few minutes. Congratulation on the collaboration. It looks like it's working pretty well. (mumbling in agreement) He's Bruno, he's Jonathan, I'm Jeff and you're watching the Cube. We're at PagerDuty Summit downtown San Francisco. Thanks for watching, we'll see you next time. (rhythmic synth music)