In Conversation with Aaron Katz, Co-Founder & CEO, ClickHouse
Ask anyone who spends time in the data ecosystem, and the name “ClickHouse” is one that has come up countless times in conversations over the last few years.
ClickHouse is a real-time OLAP (meaning, analytical) database that is known for its performance and scalability, and has a wide footprint of users around the world.
ClickHouse started its life at Yandex, the Russian search giant. It was originally created as an internal web analytics tool called Metrica, which evolved around 2009 into “Clickstream Data Warehouse” or ClickHouse for short.
The product was open sourced in 2016 and became a very popular project, with adoption at impressive scale by a number of companies including Yandex (10s of trillions of rows), Uber, Ebay, Cloudflare, Spotify, Deutsche Bank, and more.
ClickHouse was spun out into early 2021 into ClickHouse, Inc., a commercial company co-founded by Aaron Katz, Alexey Milovidov (ClickHouse’s creator), and Yury Izarilevsky (ex-Google VP Engineering), with a focus on bringing ClickHouse to all types of companies via a managed version.
ClickHouse Inc raised a $50M Series A announced in September, followed closely by a $250M Series B last month, in which my firm, FirstMark, participated.
It was a treat to welcome Aaron Katz, the Co-Founder and CEO of ClickHouse, Inc. to Data Driven NYC. Prior to co-founding ClickHouse, Aaron had extensive experience as a world-class sales leader, most recently as the Chief Revenue Officer at Elastic and the Senior Vice President of Enterprise Sales at Salesforce
Below is the video and below that, the transcript.
(As always, Data Driven NYC is a team effort – many thanks to my FirstMark colleagues Jack Cohen and Katie Chiou for co-organizing, Diego Guttierez for the video work and to Karissa Domondon for the transcript!)
[Matt Turck] I’d love to start at a high level. Maybe let’s talk about real time analytics, generally. What is the big mega trend that ClickHouse is playing into?
[Aaron Katz] I think there were a few trends that we’re seeing in the market, that we’ve observed over the past few years. What got me so excited about ClickHouse and kind of engineering the spin out of the technology, and of the team from Yandex over the past year, and partnering up with Alexey Milovidov, the creator of ClickHouse, and Yury Izrailevsky, who joined us from Google and previously at Netflix, and I think there’s a few themes. The first would be, and I anticipate a question may come in terms of how does OLAP differ from an online transactional processing database. A friend of mine commonly refers to this as the collision course between the two technologies is ensuing.
[1:02] As we emerge from the NoSQL era, over the last 10 years, the proliferation of NoSQL databases, which ironically, a lot of those NoSQL tools ultimately ended up having to write their own query engine. We’re starting to see this convergence between a traditional OLAP database management system, like ClickHouse or others, and a more traditional online transactional processing system. You could use a number of examples, let’s just use Mongo, for argument’s sake. I think the easy answer to the question could be these technologies could very easily coexist and they do today, where you could layer on top an analytics engine like ClickHouse with Mongo as the transactional record of storage.
[1:49] I think what we’re actually seeing play out in the market is the developers that are adopting these technologies aren’t drawing such a stark distinction between the two and are looking at the database that they’re selecting for a specific use case more as an endpoint, almost like an API endpoint, where they point their data to…they’re just assuming… or they don’t assume, but they’ve done the research to determine that it has the reliability, and the scalability, and the security, and all the things that you care about, whether it exists on prem, or in the cloud, or whatever deployment method that they determine. I think we’re going to see this convergence between the two.
[2:29] There’s an interesting article written by Stephen O’Grady, Stephen wrote The New Kingmakers. I don’t know, Matt, if you’ve read that book, it’s worth a read. The article’s called A Return to the General Purpose Database. I did a poor job summarizing some of the themes in that article, but he does talk quite a bit about the fact that people are now coming back to not needing all of these specialized databases for all of these bespoke use cases. That if you can have… Thank you somebody just posted that link. It’s a quick read and I enjoyed it. But I think that summarizes some of the themes that we’re seeing in the market as well. Then just this proliferation of technology, both in terms of ingestion and transformation. Obviously, things like Kafka, and DBT, and Kinesis.
[3:19] Then, on a visualization layer, things like Grafana, and Tableau, and MetaBase, and Superset, and then everything in between. I think probably the one area where we’re probably seeing less innovation is around just storage. I think storage is quickly being commoditized. The value on the stack is both on the front end and on the back end. In terms of query processing. You think about both real time analytics, which I’m sure we’re going to talk about. You think about historical analytics and predictive analytics, and I think ClickHouse falls squarely into those three analytic use cases.
[3:56] That’s a great segue, especially on the real time analytics part. Why does real time matter? In a world where, for a number of years, like people have said, “Okay, well this year is the year of real time.” Then when that didn’t happen, the next year, people we’re saying, “Oh, no, no, this year is the year of real time.” Ultimately, at least a portion of the data world felt like real time was a nice to have rather than a must have. Do you think that we are beyond that phase and firmly into the world of real time now?
[4:31] I suppose, it depends on how you define real time. I mean, if you’re talking about double digit millisecond latency on a complex query over petabytes of data, for me, that satisfies the requirement. I think we’re seeing that emerge with ClickHouse and other technologies, where you’re able to ingest petabytes of data, billions of events submitted, in the case of eBay, for example, who is adjusting literally a billion events per minute into ClickHouse, and then running very complex queries against that data set. They would, I think, argue that the analytics that they’re receiving are near real time. You could then argue the definition of that.
[5:15] That was not the case 10 years ago. We can talk about the history of ClickHouse, but when it was originally written inside Yandex, the data volumes that they were seeing were at the time some of the highest in the market, let’s say 30 billion events per day. Now, we’re seeing that being dwarfed by orders of magnitude. I think you’re starting to see the need for technologies that can both deliver the analytics experience that I just described, but also doing so in a very cost efficient way. I think there are a lot of technologies that, if you throw enough resources at, can perform at a very high degree, but costs can quickly spiral out of control. We’re seeing that as well. I think striking the combination of performance and efficiency is probably one of the more difficult feats that we’ve been hopefully able to address.
[6:09] To continue unpacking what you mentioned a minute ago, we talked about real time, but ClickHouse is also you mentioned squarely into the world of historical analysis and also predictive analytics. Can you unpack that?
[6:27] I can give you a few examples. I mentioned the Uber use case, and that’s an interesting one, and we can talk… Sorry, Ebay. Uber’s another interesting use case where they were evaluating a variety of technologies for their logging infrastructure, for their logging platform, which is being used by hundreds of developers every day for a variety of analytical workloads. The current technology presented a bit of a cost constraint and impacted developer productivity as a result. They kind of cataloged the market and chose ClickHouse because the analytics that they were running were both in real time. The actual experience, when you and I are in an Uber, and we’re going from one location to the next, from the minute we order the car to the minute it drops us off, it’s kicking off an event every few seconds back to Uber.
[7:23] They’re using that for a variety of analytical and business purposes. In doing so, they also need to report on historical data. They need to know, for example, how many cars do they need in a certain location, when to introduce surge pricing, driver and passenger sentiment, things like that. It’s not just monitoring the health of their network or traditional observability use cases in terms of monitoring logs, or metrics, or traces, but it’s more analytics that are going to drive their business.
[7:53] I think you can present a technology that offers both of those, where you can effectively report on historical data, while at the same time being able to monitor the health of your system and be able to quickly triage, for example, an issue on your network that’s causing latency with your service, and impacting your customer experience. That combination is reasonably unique. In terms of predictive analytics and modeling, I think that’s kind of the last… I wouldn’t call it the last mile, but it’s something that we’re thinking about in the future. It’s not something that we are primarily focused on today, we’re more focused on the real time analytics use cases as data is coming into ClickHouse, but we will get there over time.
[8:35] I’d love to go down memory lane a bit and go back to the origins of the company. The open-source project started at Yandex, which is the Google of Russia. I’d love for you to tell us a story.
[8:50] Sure. It first came on my radar a few years ago, the technology itself, as I started to observe its increase in popularity in the market. Earlier this year, I was introduced to Yandex, the CFO specifically, who then quickly brought in the founder and CEO of Yandex, the co-founder, Arkady Volozh. That was literally the start of the calendar year back in early January. Arkady and I started to romanticize about what it might look like to spin ClickHouse out of Yandex, as well as the core engineering team and the creator of the project, Alexey, and form a new company around this extremely popular open source database technology. I immediately reached out to two investors who I’d worked with in the past, specifically, Mike Volpi at Index Ventures and Peter Fenton at Benchmark, two people that you I believe have interviewed in the same forum in the past.
[9:46] Big fan of both of them.
[9:48] I share your sentiment. There’s a lot of very smart people that are deploying capital in this category, specifically open source infrastructure. But my belief, and I’m biased obviously, is that they are the top two, or two of the top five, let’s say. I reached out to both of them and gauged their interest in financing this endeavor. They both emphatically said yes, that they would love to work with me on this project. The four of us, Arkady, myself, Peter, and Mike engineered a spin out, and it took roughly… the gestation period was roughly nine months before we incorporated the company and determined the amount of capital that we were going to raise.
[10:27] It really took time for Alexey, and his team in Moscow to get comfortable with the idea of leaving Yandex, and joining a company that is yet to be incorporated with an unknown CEO, pre-revenue, pre-product, outside of the open source, without a customer base. That was a lot of conversations about the type of company we would build, the culture, my management style directly, their approach to engineering, the balance between open source and a business model to monetize the popularity, self-managed versus cloud, support and services, how all of that plays into building a successful enterprise software company. In doing so, I was thinking, “Gosh, me and 14 engineers in Moscow, this is a pretty big lift.”
[11:18] I thought if I could find someone, kind of the third leg of the stool, who had the same level of experience on the R&D side, product and engineering, that I had on the go to market side, this could be a really interesting combination. Honestly, the first person I thought of was Yury Izrailevsky. I first met Yury when he was running platform engineering at Netflix and led their migration to the cloud. For a number of years, I believe he was AWS’ largest customer at the time. We had dinner together five or six years ago in San Francisco. I watched as he moved to Google, and was doing amazing things within Google Cloud, and more broadly running server lists, and developer tooling.
[11:56] We took some time, but we were successful in our efforts to recruit Yury to join our company as well, and head up product and technology. The three of us formed the company, raised capital, and as you said, there was quite a bit of investor interest when we incorporated the company, and announced the financing, obviously, based off the popularity of ClickHouse and the technology itself. We pulled together a syndicate, of which you and FirstMark we’re lucky to have involved, and raised a subsequent round of financing. I’m happy to talk about the use of proceeds of capital and what our plans are, but we feel like we’ve got the right investors, a very diverse set of investors and a business plan that we’ve executed against in the past and we have a high degree of confidence we can do so again.
[12:46 Just really further the development of ClickHouse, both in terms of open source feature development, we’re going to double the size of Alexey’s team as soon as we can, as well as stand up managed service in the cloud. A true, multi-tenant, serverless cloud experience that has everything you’d expect around consumption based billing, pre-payments, credits, auto-scaling, provisioning, highly secure, fault tolerant, reliant, that supports multi-data center and geographic replication. We’ve got our work in front of us for sure, but we’re really excited to be where we are right now.
[13:26] What an incredible story. Yeah, to continue down that path, how do you think about the balance between continuing to build the open source project and the commercial product? How are you going to organize the company around that? Are you going to have different teams working on the open source project versus the commercial project? How do you plan on doing that?
[13:50] I’ve got a few thoughts. I mean, ClickHouse today is extraordinarily popular, as you know, in the open source community and developers more broadly. I think it was the second highest open source database management system on GitHub last year in terms of active contributors and perhaps the fastest growing open source database management system on GitHub alone. We are going to further feature development within the open source, obviously, licensing is an important topic that comes up frequently with open source companies.
[14:24] We’ve seen other companies in their approach to neutralize, in many cases, the threat of large cloud service providers like AWS and others. We’ve also seen companies partner very effectively with AWS, Google, Microsoft, and other cloud service providers. Look at what Grafana has been able to do or Databricks. I’m really impressed with the models that they’ve sry in terms of how can you work collaboratively with large CSPs? To answer your question directly-
[14:55] Which was going to be, while we’re on the topic, going to be my next question, what license ClickHouse under? Yeah, how do you plan on working with the hyperscale or the cloud providers?
[15:08] It’s currently governed under an Apache 2.0 license, which is extremely permissive. It allows for redistribution, modification, you can build a managed service around it. It obviously aids in growth of the project and popularity, but it comes with obvious risks that I just described. There are a variety of different licenses that we have considered that we have not yet adopted, like AGPLv3, or SSPL, which I know open source companies have deployed. We feel that today, the right decision for the community is to stay within Apache 2.0 license and we’re excited about that. Like I said before, we’re not moving away from open source.
[15:46] Quite the contrary, we’re going to double down, and double the size of the team, of the core contributors, and recruit new people into our company that understand the technology, and that have been contributing to the projects in the past. At the same time, in parallel, we are going to build a multi-tenant managed service in the cloud, which will inevitably be deployed on a variety of different cloud platforms, whether it’s AWS, GCP, Azure, whether we go to China, like I’ve done in the past, and partner with companies like Alibaba, and Tencent, and enable them to take all of the orchestration framework that we develop, some of which will be open source, some of which will be proprietary.
[16:28] We take all the tooling, all of the integrations that we develop, again, some of which we’ll release to the community, some of which we’ll rely upon the community to contribute, some of which we’ll offer ourselves. The control plane, the data plane, and potentially package all of that up eventually to where, A, if a company, a large regulated bank who for their own reasons won’t move to a public cloud service, they want to deploy it behind their own VPC or in an isolated instance, we’ll be able to support those customers in the same way that Snowflake did a very similar progression. From having a managed service to being able to offer a virtual private experience for more bespoke customers. We are thinking of following a similar path.
[17:14] Switching to products for a bit. What makes ClickHouse so special? Perhaps in contrast to other players, including open source players in the space like Druid, and Pinot, and others, and the commercial companies that are associated with those. What is the claim and the reality?
[17:45] I’ll start by saying the landscape, the competitive landscape, is very diverse, as you know. You mentioned two alternative open source projects, Druid and Pino, which recently got spun out of LinkedIn. Then, there’s a variety of cloud based data warehouse technologies. You’ve got Redshift, Bitquery, Snowflake. You’ve got some legacy technology like Vertica, Teradata, and others. I can also draw it back to this convergence that we’re predicting between OLTP and OLAP, really being served by one more general purpose database technology. I’ll also answer by saying there’s no one single feature that makes ClickHouse fast. If it was that easy, it wouldn’t have taken 10 years to develop, which is how long ClickHouse has been under development when it was first authored by Alexey 10 years ago inside Yandex.
[18:38] I will say there are a few things that we hear commonly from the community or from companies that benchmark ClickHouse. It’s obviously tied to its columnar oriented data structure, both in terms of how data is stored and how data is processed. Both of those occurring inside the column. We think about resource utilization very seriously, and compression, and how those two can aid in terms of response time, and analytics, and speed. So we invest heavily on both of those fronts, both how data is actually stored and compressed, how it’s indexed, how it’s queried, and then the amount of actual resources or CPU that you can pool and leverage to execute a single query.
[19:28] Are there any trade offs with using ClickHouse. There’s a lot of things that ClickHouse is excellent at. Are there areas where it’s not the right choice or not the right choice yet?
[19:48] Of course. There’s going to be technologies that are better suited for certain use cases. That’s another thing that we’re excited about, is the diversity of the use cases we’re seeing. We talked a bit about business analytics, we can talk about observability metrics. For example, APM. If you’re looking to deploy a hosted APM solution and you asked me objectively my recommendation, I’d say use Datadog. I’d say the product, the maturity, the experience is considerably better than trying to stitch together something using ClickHouse, for example. Now, perhaps ClickHouse could integrate with Datadog and we see that in the market as well for analytics, but for the actual instrumentation of the application, deploying the agents, I would not recommend ClickHouse today for that specific use case.
[20:36] There’s obvious exceptions where there’s going to be a better suited technology in the market. I think one thing that we’re equally excited about is the amount of integrations that we’re seeing be developed with ClickHouse. We talked a little bit about ingestion and we’re going to be investing heavily to make getting data from Kafka to ClickHouse, or Kinesis, or an integration on data transformation with DBT, a much more seamless experience than it is today.
[21:03] We talked a little bit about the front end. Today, we don’t have plans to build our own visualization layer, like you’ve seen other open source companies do. We plan to partner and support integrations for all of the major front end visualization layers, like Grafana, and Superset, and Tableau, and Looker. Companies have invested years in developing a visualization tool and then other integrations that would coexist happily with ClickHouse.
[21:36] Can you talk about the roadmap a little bit on the… I mean, obviously, you have the commercial product coming out, like you mentioned, with a lot of features, but like in terms of what the community and potential customers can expect in the next year?
[21:52] Sure, it’s broad and it applies both to open source feature development and our managed cloud service. There’s going to be quite a bit of overlap between those two, as many of the features that have been in development for years long before we created this company have been to aid in multi-tenancy, which is something that we hear a lot from users to make ClickHouse equally performant in a multi-tenant environment. Fortunately, it’s being deployed and has been deployed in that scenario at Yandex for many, many years. Our story around cross data center replication, and multi-tenancy, and security is already extremely strong. I feel very good about where we are there.
[22:38] There’s things around full disk encryption, and how do you then separate that between columnar based encryption? We’re going to invest heavily in security obviously to make sure that the largest companies in the world, as well as a lot of startups, many of whom I’m sure are represented on this call, are using ClickHouse to power the backend of their services. In some cases, I’ve seen the front end.
[22:58] I’ve seen some startups that are building web analytics applications for e-commerce or retail, for example, that show a user in milliseconds the experience that they’re having on a retail website. That’s ClickHouse powering that analytics experience. We’re going to further develop it to make sure that that experience and those companies that are betting on ClickHouse, as a business, literally to power their core business, have a highly secure and reliable experience. Then I’ve talked a bit about our investments that we’re going to make in cloud, so I don’t want to belabor that point.
[23:34] Let’s talk about go to market a little bit. I think some other people that listen to this either live or will in the video, which are looking to learn from those conversations who could benefit from this, you were the CRO of Elastic, which was a fantastic success. Maybe talk about bottoms up versus top down product led growth versus having a full on sales team. How do you plan on going about this, sequencing it towards success?
[24:06] We can spend a lot of time on that question, Matt, we may need another session. I mean, it’s something I’ve been studying for 20 years, obviously 12 years at Salesforce in distribution, and then six years at Elastic. Very different companies, very different technologies. Open source, obviously, as a technology model, we don’t need to talk about the benefits there in terms of rapid innovation. But it’s also a wonderful distribution model for a company that wants to build a commercial entity around an open source project. We’ve seen a lot of success around that. I’ve mentioned a few on this call, you can add Confluent to that list, and others, because it puts the technology in the hands of the consumers, the users, in a very frictionless way.
[24:46] Then, the business decisions need to occur between if you follow an open core strategy, and you therefore develop proprietary or enterprise features, like Elastic did, or Confluent, or others before them, or do you focus your energy on the cloud, thinking that you’re kind of building a company that will stand the test of time. I am a firm believer that the majority of workloads will shift to managed services. We’re already seeing that occur and I only think that’s going to continue.
[25:16] In terms of go to market, bottoms up versus tops down for product growth, obviously, it’s not that dissimilar from Slack. I mean, in Slack’s early days, it would distribute a very popular free product, it would hold back the enterprise features for the companies that were willing to pay for those. They built an incredible business and in doing so… You could think about a traditional enterprise sales process, my experience, yes, you need one, it’s just a function of when. The companies that we’re talking to today who are using ClickHouse, some of them I mentioned, others I haven’t. They’re out in the open, the companies that are using it. Walmart, and Microsoft, and Deutsche Bank, they expect to engage with a vendor in both a tactical way in terms of potentially how they’re using the technology, but then much more strategically, because in many ways, they’re betting their business on this company, and not just on the technology, but on the people that are going to be there to support them.
[26:18] Yes, we will build an enterprise sales motion, we’ll build an inside sales motion to capture the long tail SMB companies that are using ClickHouse, either on prem, self-managed, or in the cloud. Hopefully, the latter. Internationalization is something that we spend a lot of time thinking about, not just because… I’ll go on record to say we celebrate the Russian heritage of this software. Many will point to security vulnerabilities of software developed by Russian engineers in Russia and we don’t share that sentiment whatsoever. The team is incredibly talented, they’re super excited. They’re in the process of relocating to Amsterdam to make it just easier for all of us to do business together.
[27:04] But internationalization for open source companies is a strategic advantage. Obviously, it doesn’t know borders, in terms of adoption. It can really aid in terms of helping a company determine when they go into a new market, like Japan, or Singapore, or a major market in Europe, or Latin America, because you’ve got all these signals about how your technology is being deployed. You can make a much more calculated decision on when do I put my first salesperson on the ground? By the way, they’re not going to spend their first year trying to capture their first referenceable customer, they already have 100 companies that are using the technology already that they can start talking to.
[27:39] It’s a very warm market to enter. It shows up in the financial results of many of these companies. When I left Elastic, I believe the company was doing a little bit north of a half a billion in recurring revenue. At the time, I’m almost certain it was north of 40%, I think 41% of revenue was coming outside of the US, which is unusual for an enterprise software company at their stage. Frankly, it’s higher than even the majority of most enterprise software companies that are 10 years older than them. You get tremendous benefits in terms of distribution, the obvious ones in terms of innovation. Then drawing it back to how do you build a partner ecosystem which is accretive for everybody and not directly competitive. It really does draw, again, back to specifically around open source licensing and partnering.
[28:33] As we’re getting close to the end of the time we have, I’d love to sort of double down and sort of close with where we started, which is this convergence of OLTP and OLAP. Again, for folks OLTP being the world of transaction data and OLAP that being the world of analytical data, I think a lot of people think of ClickHouse as, okay, well this is a fantastic tool for OLAP, but more specifically real time analytics. I think what you’re saying is that the ultimate goal for the company is to be much more than that. Basically be the data store for all things. Is that a fair way of putting it?
[29:25] I point to how the market is reacting and we’re seeing this convergence today. Traditional OLAP tools where you’re running analytical workloads on top of aggregate data, there’s a lot of options in the market. It’s nothing new, it’s been around for 20 years plus. The transactional database systems that we’ve talked about in the past are very effective for, let’s say, financial transactions, credit card transactions, et cetera. We’re starting to see those two converge. I don’t think it’s just ClickHouse, so I’m not going to go on record and saying that we’re going to be the be all end all for every analytical workload, on prem or in the cloud.
[30:04] I think we’re a very attractive piece of technology for the majority of those use cases, and I described just a subset of those around observability, but we’ve got banks using us for fraud detection. There’s purpose built solutions in the market for anti-money laundering and fraud detection, specifically to financial services. You wouldn’t necessarily turn to a general purpose columnar data store like ClickHouse, but companies are. I think it’s more of a trend in the industry than it is specific to ClickHouse that you’re going to see this convergence occur.
[30:34] Again, I’ll bring it all the way back to Stephen O’Grady’s article that was posted, it’s a quick read, and it’s a good one. Also, I’ll point people to a blog that was written by a friend of mine, Martin Casado, who’s an investing partner at Andreessen Horowitz. He recently authored a blog, you should be able to find it, it talks about emerging architectures and data infrastructure. There’s some wonderful reference architectures there that talk about how all of these technologies are, to some degree, independent and to what I interpreted as a much greater degree overlapping with one another.
[31:15] Wonderful. Look, I mean, this is absolutely fascinating. Sounds like the company’s going to have a tremendous amount on its plate for the years to come, but what a formidable start with how successful the open source project is, what a dream team of co-founders, and with all the money that you need to get started, I’m very excited to see how the next years develop. Thank you for coming early into the life of ClickHouse Inc. to tell us the story. I look forward to checking in maybe a year from now, have you back, and see how the journey has progressed. We feel fortunate and privileged to get a chance to speak here with you today.
[32:12] I appreciate that, Matt. I guess one last comment, I wouldn’t be doing my job if I didn’t shamelessly inform the audience that we’re hiring engineers. For those of you who have experience in distributed systems or have an interest in what we’re doing, feel free to reach out.
[32:29] Where do people find information, clickhouse.com/careers?
[32:34] Yeah, clickhouse.com, which to Mark’s earlier point is a domain that I purchased earlier this year. When engineering the spin out, I thought, “I better buy this domain now because it’s going to get a lot more expensive when we incorporate the company and announce the financing.”
[32:48] Mark responded in the chat, “Smart man.” So you mentioned San Francisco and Amsterdam, but are you building a distributed company for folks who may be interested in applying?
[33:03] We are, yeah, we’ve got offers out to employees in Germany, Portugal, the UK, China, targeting some people in Singapore, and Australia. So absolutely, we are location agnostic.
[33:15] Wonderful, all over the globe. All right, terrific. Aaron, thank you so much. Mark, thank you as well. Really appreciate you guys sharing your stories and your journeys today. Thanks to everyone that has attended. We’ll come back to you guys shortly with the date for the next event. Thank you so much. That’s a wrap for tonight.