Today I’m happy to announce **ClickHouseInc.**, thenew home of ClickHouse. Thedevelopment team has moved from Yandex and joined ClickHouseInc. to continue building thefastest (and thegreatest) analytical database management system. Thecompany has received nearly $50M in Series A funding led by Index Ventures and Benchmark with participation by Yandex N.V. and others. Icreated ClickHouse,Inc. with two co-founders, [Yury Izrailevsky](https://www.linkedin.com/in/yuryizrailevsky/) and [Aaron Katz](https://www.linkedin.com/in/aaron-katz-5762094/). Iwill continue to lead the development of ClickHouse as Chief Technology Officer (CTO), Yury will run product and engineering, and Aaron will be CEO.
Istarted developing ClickHouse more than ten years ago, and it has never been an easy ride. Theideaof ClickHouse came up while Iwas working in Yandex as adeveloper of areal-time web analytics system. My team and Ifaced multiple dataprocessing challenges that often required custom datastructures and sophisticated algorithms, creative solutions and tradeoffs, deep understanding of domain area, hardware, and math. All these years, Ioften went to bed with endless thoughts about how we could solve yet another dataprocessing challenge. Ilove dataand processing in extreme constraints, where you have to think about bytes and nanoseconds to save petabytes and seconds. TheClickHouse team shares this passion: in my opinion, this is themain reason for ClickHouse’s success.
In 2009 we started ClickHouse as an experimental project to check thehypothesis if it's viable to generate analytical reports in real-time from non-aggregated datathat is also constantly added in real-time. It took three years to prove this hypothesis, and in 2012 ClickHouse launched in production for thefirst time. Unlike custom datastructures used before, ClickHouse was applicable more generally to work as adatabase management system. After several years Ifound that most departments in my company were using ClickHouse, and it made me wonder: maybe ClickHouse is too good to run only inside Yandex? Then we released it in [open source](https://github.com/ClickHouse/ClickHouse) in 2016.
Making ClickHouse open source was also not an easy decision, but now Isee: doing open source is hard, but it is abig win. While it takes atremendous effort and responsibility to maintain apopular open-source product, for us, thebenefits outweigh all thecosts. Since we published ClickHouse, it has been deployed in production in thousands of companies across theglobe for awide range of use cases, from agriculture to self-driving cars. In 2019 we spent over athird of our time abroad organizing various ClickHouse events and speaking at external conferences, and we’re thrilled to see you all again in person once travel restrictions become less severe. Thefeedback and contributions from our community are priceless, and we improve thequality of implementation, thefeature completeness, and making product decisions with thehelp of our community. One of our main focuses is to make ClickHouse welcoming for contributors by making thesource code easy to read and understand, with theprocesses easy to follow. For me, ClickHouse is ashowcase so everyone can learn theideas in dataprocessing.
Ilike to present ClickHouse as theanswer to many questions in software engineering. What is better: vectorization or JIT compilation? Look at ClickHouse; it is using both. How to write thecode in modern C++ in asafe way? Ok, look at thetesting infrastructure in ClickHouse. How to optimize thememcpy function? What is thefastest way to transform aUnix timestamp to date in acustom timezone? Ican do multiple-hour talks about these topics, and thanks to theopen-source, everyone can read thecode, run ClickHouse and validate our claims.
Themost notable advantage of ClickHouse is its extremely high query processing speed and datastorage efficiency. What is unique about ClickHouse performance? It is difficult to answer because there is no single "[silver bullet](https://www.youtube.com/watch?v=ZOZQCQEtrz8)". Themain advantage is attention to details of themost extreme production workloads. We develop ClickHouse from practical needs. It has been created to solve theneeds of Metrica, one of the[most widespread](https://w3techs.com/technologies/overview/traffic_analysis) web analytics services in theworld. So ClickHouse is capable of processing 100+ PBs of datawith more than 100 billion records inserted every day. One of theearly adopters, Cloudflare, uses ClickHouse to process alarge portion of all HTTP traffic on theinternet with 10+ million records per second. As ClickHouse developers, we don’t consider thetask solved if there is room for performance improvement.
Query processing performance is not only about speed. It opens new possibilities. In previous generation datawarehouses, you cannot run interactive queries without pre-aggregation; or you cannot insert new datain real time while serving interactive queries; or you cannot just store all your data. With ClickHouse, you can keep all records as long as you need and make interactive real-time reporting across thedata. Before using ClickHouse, it was difficult to imagine that analytical dataprocessing could be so easy and efficient: there is no need for adozen pre-aggregating and tiering services (e.g. Druid), no need to place huge datavolumes in RAM (e.g. Elastic), and no need to maintain daily/hourly/minutely tables (e.g. Hadoop, Spark).
Most other database management systems don’t even permit benchmarks (through theinfamous "DeWitt clause"). But we don’t fear benchmarks; we [collect them](https://github.com/ClickHouse/ClickHouse/issues/22398). ClickHouse documentation has [links](/docs/en/getting-started/example-datasets/) to publicly available datasets up to multiple terabytes in size from various domain areas. We encourage you to try ClickHouse, do some experiments on your workload, and find ClickHouse faster than others. And if not, we encourage you to publish thebenchmark, and we will make ClickHouse better!
Yandex N.V. is thelargest internet company in Europe and employs over 14,000 people. They develop search, advertisement, and e-commerce services, ride tech and food tech solutions, self-driving cars... and also ClickHouse with ateam of 15 engineers. It is hard to believe that we have managed to build aworld-class leading analytical DBMS with such asmall team while leveraging theglobal community. While this was barely enough to keep up with thedevelopment of theopen-source product, everyone understands that thepotential of ClickHouse technology highly outgrows such asmall team.
We decided to unite theresources: take theteam of core ClickHouse developers, bring in aworld-class business team led by [Aaron Katz](https://www.linkedin.com/in/aaron-katz-5762094/) and acloud engineering team led by [Yury Izrailevsky](https://www.linkedin.com/in/yuryizrailevsky/), keep thepower of open source, add theinvestment from theleading VC funds, and make an international company 100% focused on ClickHouse. I’m thrilled to announce ClickHouse,Inc.
Companies love ClickHouse because it gives tremendous improvements in dataprocessing efficiency. But it is mostly about thecore technology, thedatabase server itself. We want to make ClickHouse suitable for all kinds of companies and enterprises, not just tech-savvy internet companies who are fine with managing their clusters. We want to lower thelearning curve, make ClickHouse compliant with enterprise standards, make ClickHouse service to be instantly available in thecloud in aserverless way, make auto-scaling easy, and much more.
Our mission is to make ClickHouse thefirst choice of analytical database management systems. Whenever you think about dataanalytics, ClickHouse should be theobvious preferred solution. Isee how many companies already benefit from ClickHouse and I'm very eager to make it even more widespread and universally accepted across theworld. Now we have thebest engineers and thebest entrepreneurs together and we are ready for themission.
