 I think generative AI will be absolutely an integral part of probably every observability product. It's a nice abstraction. It makes it easier. It's a more natural way to interface with these tools. So I think it's going to be a productivity benefit for users. Hi, this is your host, Sapin Paltia, and welcome to your foreign newsroom. And today we have with us Jeremy Burton, CEO of Observe. Jeremy is great to have you on the show. Great to be here. Thanks for the invite. Yeah, it's my pleasure to host you here today. Today, we are basically going to talk about Observe Hubble announcement that you folks made earlier this month. But before we go and talk about that, especially since you are here on the show, I would love to hear tell us about the company and its role in this modern cloud-centric, cloud-driven, clouded world. Yeah, the company was founded back in late 2017. So it's been quite a long journey. But our observation was, no pun intended, was that cloud-native apps will become more complicated because of microservice architectures. And when things went wrong, it was possibly a brand-damaging event because so many companies were moving to produce a digital experience. IT was no longer part of the business. IT was the business. And so when the IT systems go wrong, you can't do business with your customers. And so we thought, well, given that most people now are developing a core competence of software and are delivering new releases daily, it's going to become a big deal to troubleshoot those problems when those applications go wrong. So Observe really took a fresh approach. We tried to bring together what in the past has been sort of three discrete markets. Log analytics, which maybe someone like Splunk was the leader. Monitoring, which maybe someone like Datadog was the leader. And APM, which I don't know, maybe someone like New Reliquis was the leader. So we're trying to produce one product that does all three because to troubleshoot these modern applications, it's not enough to just look at the logs or look at the metrics or look at the traces. You want to be able to look at any and all data that can help you troubleshoot that problem in one view. So that's really what we're trying to do. As I said, it's been about a five-year journey so far. When we look at among all the things that developers are responsible for, we look at observability. Talk about how we have seen the evolution of observability, tracing, monitoring. Also, we saw a lot of projects that consolidated open telemetry or percents from Google and LightStep. The whole point is that these things are moving into developers' pipeline. So talk a bit about how does observability fit in developers' workflow and when they invest their time and resources, what do they get out of it? It's a great question. Developers want productivity. They want to spend the maximum amount of time possible writing code, building new features, creating better experiences for their customers. I think that's what everybody is excited about. Obviously, bugs or worse, incidents detract from that. What we've seen, I think the independent data I looked at, that the meantime to resolve an incident in your average company is 172 minutes is the last data I saw. It's almost three hours. Now, best in class is actually 43 minutes. I'm sure if you've got Google's SRE team, you can resolve an incident in 43 minutes. The problem is not everybody has Google's SRE team. I view observability as a way to try and allow almost every team to drive down their meantime to resolve incidents from almost three hours down to that best in class 45 minutes. Software should be able to deliver that productivity because then that two hours that you're saving is going to be spent by a software developer doing what they love, your building features. Nobody loves being woken up in the middle of the night working on an incident. Observability done right will minimize the time required to resolve those incidents and as a side effect allow folks to spend more time doing what they love. Yes, while it's true that we should free developers' time and doing what they love doing also for a business perspective, they should be focusing on writing applications that add value to their business. That's where companies like Observe come to picture because of course, there's a fact that open source is driving the bird, but open source can solve different problems. Companies need additional features, they need functionality, they need a toe to choke. Talk about the role that Observe plays in making things easier for developer lures the barrier of entry so folks can also add observability into their pipelines. You bring up a great point about open source and there are many people build applications on open source frameworks. Our view is that look, observability, you could assemble an observability stack from open source, but I don't believe that it's anybody's core competence to build an observability stack. An observability stack is a means to an end, it's a way for you to resolve incidents. We offer a SaaS service and we manage everything. An engineer coming into Observe, I think they're just trying to get through their day. They're just trying to find the root cause of an issue, do the investigation and go back to what they love doing. I feel like observability is one of those areas where although there are open source stacks, I think most people want to buy an outcome. Give me a service, make sure it's always available, make sure it's easy to get going with and make sure it works. My guys who are using it can get in and out and get back to doing their day job. I think that's been our mantra and our approach. Let's make it seamless for the user. We manage everything from collecting the telemetry data right the way through to when the user queries it and we guarantee the availability and the latency and all that good stuff so they don't have to worry about it. It should just work. Can you also talk about the evolution of Observe as a company, your services, your offering which can also include the latest Hubble announcement so that you are also kind of evolving to be with your customers wherever they are in their journey. We offer a SaaS software service so we don't deploy the service on-premises. It's only in the cloud and the latest incarnation, the Hubble release is really a culmination of work that was created by our first 50 or so customers. When you're a startup, you get going with your initial product, you get feedback, you iterate and I think this release that we pushed out a couple of weeks ago was really designed to get the scale and the latency and the broad feature set to where most of our customers would want it. We think scaling the platform to a petabyte of data a day, that will make it enterprise ready. Get in latency of data down below 15 seconds so from the moment the data is created to the moment you can query it, it's 15 seconds. The user interface, as I was saying earlier, it's got to be super simple, it's got to be super easy to use. I think generative AI is going to add a lot in terms of making it very, very straightforward for new users to get going. I think most companies would complain about how hard it is to hire. The skills that they would need to troubleshoot an application are few and far between. Generative AI should make that on-ramp a whole lot easier than it's ever been. The product I think from in its early days we didn't have generative AI. I think it was much harder for new users to get going. Today we've got generative AI in the Hubble release. I think it's a lot easier for users to get going. We didn't have the scalability when we first shipped the product. I think the first time we actually got a customer back in 2020 we were concerned that we could do one terabyte a day. Now we can do a petabyte a day, just three years later. I think every startup is on a journey. You always try and respond to the feedback that comes from your prospective customer. Then look, there's a sort of narrow window of time you've got to build those features. You want to make sure that you build the features before you run out of money because then you can get more customers, you can raise more money and the journey continues. You talked about generative AI. I was going to ask a question about generative AI. I want to look at it from a different perspective. One is generative AI workloads and second is generative AI for observability. How do you look at it? Both ways. I think generative AI will be absolutely an integral part of probably every observability product. It's a nice abstraction. It makes it easier. It's a more natural way to interface with these tools. I think it's going to be a productivity benefit for users. I also think what we're starting to see is a new class of applications that are built against either open AI directly or private, large language models. People are going to want to observe that interaction. Is my LLM hallucinating? What's the latency? What's the cost? There's a whole bunch of metrics and I think telemetry that you're going to want to look at in your generative AI application and the observability tools should provide visibility into that as well. Let's also talk about series A funding. Talk a bit about how the company has grown over the years and what are the areas where we will be investing with this funding ground. We did our series A about three years ago. That was $15 million. For the last three years, we've actually ran the company off of debt financing. It's a convertible note underwritten by Hill Ventures that did our series A as well. It's a little bit unusual because people normally talk about debt financing as a bridge. It's like, oh, we raised the series A and we wanted to wait a little bit to maybe get some better numbers before we went and did the series B so we'll do some debt. For us, it's been a deliberate strategy to defer the company dilution to a point where we've got a real and functioning business. We've got real revenue, real customers and the company therefore is more valuable. When we do the series B, our debt will convert to equity and we'll then start to look like probably any normal startup, but we're a little bit unusual. A lot of that is down to Hill Ventures. I think they've been very good with the team here. Helping us defer the dilution until the company was more valuable because the problem that we're trying to solve, it's taken us five years. A lot of founding teams would get washed out. If we did a series B a year after we did our series A, then we would have got a terrible valuation. We would have had to give up most of the company and it would have been very demotivating. The debt allows us to defer that dilution. It's been quite unusual, but I think quite satisfying because although we will have to raise probably money next year, we've got a very good business at this point and so we should be able to get a decent valuation. Maybe unusual for us, it took us a while to get our public free trial going. We spent a lot of time trying to prove out the core value of the product and then we had to go back and work on the user interface and the easy on-ramp and a product led growth strategy. We're pretty excited that going to KubeCon is everyone who's interested who comes to our booth, we can now direct them to an observed free trial. I know that's not a new thing for every vendor, but it's certainly new for us and we're pretty excited about that. We've also got some interest in announcements coming with Snowflake. Unlike any other observability vendor, our data store is Snowflake. A number of our customers who have telemetry data in Snowflake, they would like us to share that data out to other Snowflake users so that that telemetry data can be combined with other business data to give new insights. For example, did any of my best customers have an error on my website? Typically, best customers is stored in some financial system and errors on the website is stored obviously in the web server logs. If you can join those two together, you can get business context for telemetry. We're pretty excited about that. We also are going to allow Snowflake customers to use their existing Snowflake contract to run observed. 99% of our customers today, they just buy observe and Snowflake is embedded, but obviously enterprise customers, they've got their own relationship with Snowflake and if they want to use their Snowflake contract to run observed, they're going to be able to do that. There's a couple of interesting things with Snowflake that we're going to be talking about. Jeremy, thank you so much for taking time out today and talk about this new announcement, the whole funding, and also more importantly, the whole observability landscape, how it's evolving. Thanks for all those insights and I'd love to chat with you folks again. Thank you. Great. Thanks for the opportunity.