"Change the habits of humans using data", five-year-old Kyligence made a new wish

Changing is never easy, but Kyligence decided to give it a try.

At the Data & Cloud Summit 2021 held on July 30, Kyligence co-founder and CEO Han Qing announced Kyligence's new vision-" change the habits of human use of data ". He said that the future-oriented enterprise-level data service and management platform should allow data to find people who need it, instead of people looking for data, through various new technologies and new architectures such as artificial intelligence, voice interaction, intelligent recommendation, and knowledge graphs. , And further make data serve people.

In addition, Kyligence announced the brand new "intelligent data cloud" strategy, which will build the next generation of AI-enhanced data service and management products based on artificial intelligence, cloud native and other technologies; and released the latest product Kyligence v4.5, aiming to adopt A more powerful, more efficient, and more convenient data intelligence platform empowers enterprises to transform into digital.

Intelligence + data + cloud, Kyligence's strategy is fully upgraded

Data warehouse technology was born for decades. But in the cloud era, related industry assumptions have changed. Han Qing, co-founder and CEO of Kyligence, elaborated on the following points:

1. The data management model changes from centralized to distributed;
2. The use of data has changed from a small number of decision makers and experts to front-line business personnel and ordinary workers;
3. The way of data consumption ranges from finding answers to known problems to providing advance insights into unknown problems through smart recommendations and other functions.

Adapt to changes in order to control changes.

Faced with the difficulties caused by massive data, multiple cloud platforms, complex data sources, integration between technologies, and integration between platforms, how should enterprises manage and analyze data well? How to find valuable data?

Kyligence officially announced a comprehensive upgrade of its strategy- committed to building an "Intelligent Data Cloud" (Intelligent Data Cloud) platform for enterprise customers: to enhance data management capabilities on the basis of stronger analysis capabilities, and to further replace manual work with artificial intelligence. Native further replaces the Hadoop-based infrastructure, allowing data services and management to play a core role, helping enterprises to intelligently manage the most valuable data, and supporting the comprehensive digital transformation of enterprises.

"Intelligent data cloud is not only Kyligence's business practice in the global market in recent years, but also a reflection and summary of technology development trends in the cloud-native era," Han Qing said, "The transformation of data warehouses in the cloud era has just begun. , New technical architectures, new usage methods, and new cost structures will profoundly change this industry. Human use data in the future should be as simple and convenient as using cloud computing today. You only need to pay attention to the data itself, not to the end. On which platform, the data can be accessed on demand. ”

Kyligence currently supports Microsoft Azure, Amazon AWS and Huawei Cloud platforms, and is actively deploying other public cloud platforms. In addition, with the rising demand for private cloud architecture by enterprises, Kyligence officially launched the Xuanwu plan to accelerate the implementation of next-generation private cloud products based on Kubernetes and distributed object storage architectures. Kyligence will provide a private cloud environment for large enterprise customers Run AI to enhance the ability of data service and management, and hope that more partners will join and build together.

In addition, Kyligence also announced the Kyligence Partner Network partner program, which will empower partners from training certification, resource support, and promotion cooperation, and work together to bring better services to customers around the world.

Kyligence new version released: full-scenario OLAP and AI are the focus

Kyligence released a new version v4.5, which focuses on the theme of full-scenario OLAP, integrates various technological innovations and breakthroughs, and provides users with easy-to-use, high-performance, and high-concurrency AI enhancements through the use of machine learning and artificial intelligence technologies. Data service and management platform to improve the efficiency of data engineering.

Kyligence 4.5 version has the following main features:

Organic integration of Apache Kylin and ClickHouse

The integration of the technical advantages of Apache Kylin and ClickHouse is the core of the full-scene OLAP in the latest version of Kyligence. Through Kyligence Smart Tiered Storage™️ technology, ClickHouse is organically integrated into the base of Kyligence products and analyzed in the original aggregation On top of the high performance, it has effectively improved the performance and advantages of detailed analysis, Ad-Hoc query and other scenarios. Kyligence can provide users with comprehensive OLAP service capabilities, and even further provide commercial support for ClickHouse. This feature is available in both the enterprise version (Kyligence Enterprise) and the public cloud version (Kyligence Cloud).

Batch flow integration

The new version of Kyligence officially supports batch and stream integration capabilities, which further expands the full-scenario OLAP capabilities. Through only one data model and one SQL statement, batch data and stream data can be accessed at the same time, providing a unified query interface for data applications. Help enterprises to extremely simplify the data application architecture, use the same system and architecture to meet different needs at the same time, so as to respond to business agility faster.

AI enhancement engine

Based on the AI-enhanced engine, Kyligence can automatically recommend data models based on business analysis behaviors, helping companies to identify and accumulate data assets from massive analysis loads, and intelligently update models according to business changes to achieve automated construction and management. In addition, the AI enhancement engine can also automatically clean up inefficient storage and continuously optimize TCO.

Enterprise-level operation and maintenance management

At the same time, the new version helps enterprises realize multi-tenant deployment and management through a complete enterprise-level operation and maintenance management system, and realizes automated production operation and maintenance through indicator monitoring, alarms, etc., to meet the strict IT compliance requirements of banking, insurance and other industries .

Support multiple cloud platforms

Kyligence now supports public cloud platforms such as Microsoft Cloud Azure, Amazon Cloud AWS, and Huawei Cloud. With the continuous maturity of cloud-native technology and the rising demand for private cloud architecture by enterprises, Kyligence will further accelerate the pace of cloud-native deployment, give full play to the existing technology accumulation, and empower enterprises to accelerate the realization of data cloud.

Want to change the habits of humans using data, where is Kyligence's confidence?

Strategic upgrades and new visions are inseparable from the support of strength, and Kyligence's confidence is inseparable from "open source".

Kyligence was founded by the core team of Apache Kylin (the leading open source distributed OLAP analysis engine), and Apache Kylin is the first Apache open source project led by Chinese.

In October 2014, Apache Kylin was open sourced and became an Apache incubator project in November of that year. One year later, the Apache Foundation approved Apache Kylin to officially graduate as a top-level Apache project, and the Kylin project was highly recognized.

In March 2016, Kylin's core team founded Kyligence. At the beginning of the company's establishment, Kyligence co-founder and CEO Han Qing explained the original intention of establishing a commercial company:

We found that the industry needs such a solution that can solve the problem well. There are a lot of users seeking such commercial support, and we are also optimistic about the future of big data in China. On the other hand, behind every successful open source project there will be a startup company to promote, because startup companies will be more flexible and can do things better. In addition, we also hope that in the entire big data industry, a startup company that specializes in the underlying technology can make some breakthroughs and do something in the industry. These are the original intentions when we founded Kyligence.

Kyligence utilizes the transformative power of OLAP to big data, based on Apache Kylin, provides a set of enterprise-level big data analysis solutions for private cloud, public cloud and hybrid cloud.

However, if you want to further improve the efficiency of data use and conform to the technological development trend of the artificial intelligence era, Kyligence also needs other support, such as a new generation of open source data and AI processing language MLSQL. In March of this year, MLSQL author Zhu Hailin joined Kyligence in an attempt to build a one-stop data analysis and AI platform.

MLSQL is a standard big data/machine learning language. At the usage level, MLSQL can achieve to solve the problems of Data+AI in one language, and open the entire Data+AI pipeline to ; at the bottom level, MLSQL avoids overly complex systems and chooses to unify at the engine level.

MLSQL author Zhu Hailin said that one language, one engine, and one storage are enough. These three components can theoretically solve most problems, and the maintenance cost is extremely low, and the threshold for use is low. For example, many products and operations can directly use MLSQL, and a little knowledge of SQL can be fully mastered.

So, in what ways can MLSQL be combined with Apache Kylin to help Kyligence's products and services?

Zhu Hailin said that big data consists of BI, AI, ETL (batch streaming) and Ad-Hoc query. Kirin has covered BI + part of the data processing, that is, Ad-Hoc query, but it is currently in AI and comparison. The flexible ETL has no coverage very well. and MLSQL is relatively strong in the field of ETL and AI, so it plays a very good supplementary role. Combining Kirin and MLSQL can cover the complete big data field and AI field scenarios. Currently, Kylin and MLSQL have made some combination at the open source level: MLSQL is used as the front of Kylin, allowing users to aggregate data from various data sources into hive, and then use Kylin for accelerated query, so that users have Package solution.

Kylin pursues the ultimate performance of the machine, can achieve up to a trillion rows of data in millisecond to sub-second query latency, and supports high concurrency. MLSQL's long-standing practice is to lower the threshold for users, let more people use the data, and take advantage of the value of the data. What impact will such a powerful combination of two open source projects have on work efficiency and data processing methods? We will wait and see.