Aiming to speed up the accessibility of companies’ vast data troves, data warehouse virtualization company AtScale Inc. today updated its flagship Adaptive Analytics platform.

AtScale’s platform serves as an abstraction layer for Hadoop clusters and other back-end data stores such as Snowflake, Google BigQuery, Amazon Redshift and Microsoft Azure SQL Data Warehouse. The technology makes these data stores more accessible to a range of business intelligence applications such as Excel spreadsheets and Tableau Software Inc.’s visualization tools, without the need for complicated “extract, transform and load” procedures.

Essentially, what AtScale does is make it easy for companies to tap into their data lakes and warehouses, which are repositories of structured and unstructured data typically used for business analysis, and gather insights from this information. With AtScale’s Adaptive Analytics platform, companies can build analytics models that connect any supported BI tool to any of the above data stores.

AtScale also boosts the performance of these tools with its Adaptive Cache query acceleration technology, which is able to analyze query patterns in real time to optimize responses. It works by intercepting queries and rewriting them to a common metadata repository that includes a semantic model, governance rules and a data lineage map. The Adaptive Cache creates summary tables, and machine learning algorithms generate “smart aggregations” that anticipate future queries based upon historical activity.

AtScale Chief Executive Officer Chris Lynch (pictured) spoke more about how the company’s platform makes data accessible to business intelligence during an interview on theCUBE, SiliconANGLE’s mobile livestreaming studio, last August:

The new Adaptive Analytics 2020.1 release introduces a new “multisource intelligent data model” that enables users to create logical data models without copying or transforming their existing data structures. That helps accelerate query times by assembling the data that’s needed in a “just-in-time” fashion, and by maintaining acceleration structures for subsequent workloads, the company said.

The new release also features “self-optimizing query acceleration structures” that work by incorporating additional data into the creation of acceleration structures. This helps to alleviate the “lowest common denominator” approach to query planning that involves lots of manual data provisioning and movement.

“AtScale’s Autonomous Data Engineering automatically determines the necessary structures and their optimal location,” the company said.

Another highlight is the new “virtual cube catalog” feature that’s said to speed up discovery times through new data lineage and metadata search capabilities that can integrate natively into existing data catalogs.

With the new features, AtScale said, its latest benchmark tests show it’s able to increase query performance by up to 10 times, user concurrency by 60 times and reduce costs by 10 times.

“AtScale’s latest release makes it easier to achieve both scale and performance for critical big data analytics by automating the data engineering tasks across today’s hybrid cloud and multicloud data platforms,” said David Menninger, senior vice president and research director at Ventana Research.

Photo: SiliconANGLE