Kyligence

Definition

Kyligence is a software company founded by the original creators of Apache Kylin, an open-source distributed analytics engine for large-scale OLAP queries, and it builds a commercial enterprise platform extending Kylin's core multidimensional cube technology with cloud-native deployment, automated cube design, and enterprise management tooling. The company's product targets organizations that need sub-second analytical query response over big-data-scale datasets without hand-tuning every OLAP cube.

Overview

Kyligence exists to solve a problem that the open-source Apache Kylin project addressed at a technical level but that many enterprises struggled to operate at scale: building and maintaining multidimensional OLAP cubes over big-data platforms like Hadoop and Spark to deliver interactive query response times, a process that historically required significant manual modeling and tuning expertise. The company was founded by Kylin's original creators, who had built the project at eBay and later contributed it to the Apache Software Foundation, giving Kyligence deep first-party knowledge of the underlying engine. Mechanically, the core technology precomputes and stores aggregated results across combinations of dimensions and measures, similar in spirit to traditional OLAP cubes, so that a query touching a subset of those precomputed combinations can be answered from stored aggregates rather than scanning raw data at query time, delivering sub-second responses even over datasets with billions of rows. Kyligence's commercial platform adds automation on top of this, using workload analysis to recommend which cube combinations to build rather than requiring a data engineer to manually specify every dimension combination, along with cloud-native deployment options that separate storage and compute for elastic scaling. Among big-data OLAP and analytics platforms, Kyligence's technology is closest in spirit to Apache Druid, another engine built for fast aggregation queries over large event-style datasets, though Kylin's cube-precomputation model differs mechanically from Druid's real-time ingestion and indexing approach, and both differ from GPU-accelerated engines like SQream that avoid precomputation in favor of raw scan speed. Kyligence's commercial value proposition is analogous to how Cloudera or Databricks commercialize Hadoop or Spark: taking a complex open-source engine and packaging it with automation, support, and enterprise tooling that make it operable by teams without deep in-house expertise in the underlying project. In practice, organizations use Kyligence's platform to power interactive BI dashboards over datasets too large for traditional data warehouses to serve instantly, connecting standard BI tools like Tableau or Power BI to the precomputed cube layer so that business users get sub-second response times on ad hoc slicing and dicing of large datasets. The trade-offs of the cube-precomputation approach used by both Kylin and Kyligence include storage overhead, since precomputed aggregates across many dimension combinations consume significant additional space, and less flexibility for entirely ad hoc queries that touch dimension combinations not covered by any precomputed cube, which fall back to slower raw scans. Teams whose query patterns are highly unpredictable or exploratory may find engines built for flexible raw-scan performance, like ClickHouse or Druid, a better fit than a cube-based architecture optimized around known, recurring query patterns.

Key Concepts

• Founded by the original creators of Apache Kylin
• Extends Kylin's OLAP cube technology with cloud-native deployment
• Automates cube design using workload analysis instead of manual tuning
• Precomputes aggregates across dimension combinations for fast queries
• Separates storage and compute for elastic cloud scaling
• Integrates with standard BI tools like Tableau and Power BI
• Targets sub-second query response over big-data-scale datasets
• Provides enterprise management and support around the Kylin engine

Use Cases

Frequently Asked Questions

Related in Database