Databricks, provider of the Lakehouse data and AI platform, has extended the platform’s capabilities with the addition of advanced data warehousing and governance, data sharing innovations including an analytics marketplace and data clean rooms for data collaboration, automatic cost optimisation for ETL operations, and machine learning (ML) lifecycle improvements.
The company, founded by the creators of open source solutions Delta Lake, Apache Spark and MLflow, works across business sectors including financial services, where its customer base includes the likes of Nasdaq, ABN Amro, Schroders, FIS, and Swedbank.
“Our customers want to be able to do business intelligence, AI and machine learning on one platform, where their data already resides. Databricks Lakehouse Platform gives data teams all of this on a simple, open, and multi-cloud platform,” says Ali Ghodsi, co-founder and CEO at Databricks.
The company’s additional data warehousing capabilities include Databricks SQL Serverless, available in preview on AWS and providing fully managed elastic compute for improved performance at a lower cost; Photon, a query engine for lakehouse systems that will be made generally available on Databricks Workspaces in coming weeks; open source connectors for Go, Node.js, and Python, to make it simpler to access the lakehouse from operational applications; and Databricks SQL CLI, enabling developers and analysts to run queries directly from their local computers.
Data governance additions include Unity Catalog, which will be made generally available on AWS and Azure, and provides centralised governance for all data and AI assets, with built-in search and discovery, and automated lineage for all workloads.
The company’s marketplace for data and AI will be available later this year, providing a place to package and distribute data and analytics assets. Unlike pure data marketplaces, Databricks’ offering enables data providers to package and monetise assets such as data tables, files, machine learning models, notebooks and analytics dashboards. Cleanrooms, also available later this year, will provide a way to share and join data across organisations with a secure, hosted environment and no data replication required.
ML advancements include MLflow 2.0, which includes MLflow Pipelines that can handle the operational set up of ML for users. Instead of setting up orchestration of notebooks, users can define the elements of the pipeline in a configuration file and MLflow Pipelines manages execution automatically. Beyond MLflow, Databricks has added serverless model endpoints to directly support production model hosting, as well as model monitoring dashboards to analyse real-world model performance.
Delta Live Tables is an ETL framework using a simple, declarative approach to building data pipelines. Since its introduction earlier this year, Databricks has expanded the framework with a new performance optimisation layer designed to speed up execution and reduce the costs of ETL.
Ghodsi concludes: “These new capabilities are advancing our Lakehouse vision to make it faster and easier than ever before to maximise the value of data, both within and across companies.”
Subscribe to our newsletter