Looker, the data platform builder looking to expand the reach of its “horizontal” foundation, added another component this week with the integration of its tools with the beta release of Google’s cloud-based BigQuery ML.
The combination is aimed at speeding results derived from data science workflows and would allow users to act on data insights with predictive metrics. The integration also would automate some processes so data teams could create machine learning models in BigQuery using the Looker data platform tools.
Among them is Looker Blocks. The company said Wednesday (July 25) that BigQuery ML’s predictive functionality is being integrated into new and existing versions of Blocks, thereby allowing users to generate predictive metrics in applications or on dashboards.
Lloyd Tabb, Looker’s co-founder and CTO, contends that “much of the work in machine learning centers around data preparation and ML model evaluation and tuning.” The integration with BigQuery ML means the company’s tools handle data preparation while the Google tool “does the learning,” Tabb added.
The integration also means Looker Blocks can be used to tune machine learning models to combine predictions into data workflows, the partners said.
The integration is part of a broader Google effort to bring machine learning to its flagship BigQuery data warehouse. The intent is to add machine learning features to help ease the process of moving large data sets in and out of data warehouses. The goal is to reduce the amount of data movement as developers build and refine models.
“BigQuery ML brings machine learning closer to where customers are storing large datasets, so they can quickly create and deploy models, at scale,” said Sudhir Hasbe, Google Cloud’s director of product management.
For Looker, Santa Cruz, Calif., integration with BigQuery ML provides another tool running on its expanding data platform aimed at reducing data prep hassles so developers can focus on “operationaliz[ing] the outputs of ML models….”
The seven-year-old company’s analytics and machine learning tools form the basis of its data platform along with business intelligence and visualization capabilities. Launched in 2016, Looker Block was designed to allow developers to use pre-built data sets and visualizations. Looker provides a handful of blocks, such as weather and demographic data, while others are provided by the company’s partner network.
Google (NASDAQ: GOOGL) announced the beta release of BigQuery ML this week during the company’s cloud application event. It allows developers to create and run machine learning models in BigQuery using standard SQL queries. “BigQuery ML increases development speed by eliminating the need to move data,” Google said.