Kyligence Launches Cloud-Native Edition Of Its Big Data Platform
New AI and automated data modeling capabilities speed up data preparation and query performance – even when working with petabytes of data.
Big data analytics software developer Kyligence this week unveiled Kyligence Cloud 4, the first cloud-native release of the company’s AI-enhanced analytics platform that can deliver sub-second response time against petabytes of data.
The company is the latest of a number of database and data analytics companies to offer their software through the cloud. Kyligence Cloud 4 is available on the Amazon Web Services and Microsoft Azure platforms.
“The whole point is sub-second interactive queries against very large datasets. It’s about performance for analytics in the cloud,” said George Demarest, Kyligence head of marketing, speaking of the company’s new software in an interview with CRN.
[Related: 10 Hot Big Data Companies You Should Watch In 2021]
The Kyligence system is based on Apache Kylin, an open-source, distributed analytics engine for performing multi-dimensional analysis on huge datasets. The original Kylin technology was developed by the founders of Kyligence (“Kylin plus intelligence”) who founded the company to provide a commercial version of the technology with added capabilities and services.
Kyligence, with headquarters in San Jose, Calif., and Shanghai, China, was launched in 2016. The company has raised $48 million in funding.
A key selling point of the Kyligence system is its OLAP functionality that pre-aggregates data in multi-dimensional indexes or “cubes,” greatly speeding up queries and analysis of data. The system can handle datasets in cloud-based data warehouses and data lakes with hundreds of terabytes and even multiple petabytes of data.
While Kyligence is competing with cloud data warehouse service providers such as Snowflake and AWS, customers are also using the company’s software in conjunction with those platforms to pre-aggregate data to cut down on billable computation time with those vendors, said Li Kang, head of Kyligence, North America, in the CRN interview.
“The more you run those queries, the more money you save,” Li said.
Earlier versions of Kyligence were closely tied to the Hadoop big data platform. But Demarest said later editions, culminating with the Cloud 4 release, have reduced that dependence as Hadoop has become less popular for big data tasks.
The new Kyligence Cloud 4 offers the elasticity that comes with a cloud-native architecture and the separate scaling of compute and storage functions, Li said. It also allows Kyligence to take advantage of cloud clusters and cloud object storage systems.
New in the Cloud 4 release is an AI-augmented engine that uses machine learning algorithms and auto-indexing, leveraging query history and previous user behavior to continually improve performance.
The release also offers data modeling automation to reduce data preparation time for data science and data analysis workloads. Demarest said the new capabilities can reduce data preparation from days or weeks to hours or even minutes.
A new unified semantic service in Cloud 4 creates a single, consolidated view across data sources that can be easily accessed by SQL business analysis tools like Tableau, Excel or custom-built SQL applications. (Kyligence provides its own lightweight data analysis tool for accessing data in the Kyligence system.)
And new “Smart Pushdown” functionality provides intelligence query routing to improve query performance, even for ad hoc or detailed queries, and eliminates the need to move data for data discovery and exploration.
Kyligence sells both direct and through channel resellers and works with systems integrators (including global systems integrator Cognizant) and other implementation partners. Technology partners include cloud and data management platform vendors and business analysis tool and application developers.