Looker Sees A Way To Boost Data Scientist Productivity With New Data Workflow Capabilities

Data analytics platform developer Looker is making it easier for users to tap into data science tools that help reduce time-consuming data preparation chores and allow data scientists to focus on more valuable model-building tasks.

Tuesday Looker unveiled a new release of Action Hub, the Looker platform component that contains pre-built functions, with capabilities that improve and optimize data science workflows.

"Cleaning and preparing data is not the most valuable use of time for data scientists," said Looker CEO Frank Bien in a statement.

[Related: The 2018 Big Data 100]

Looker is one of the new generation of business intelligence and data analysis software that's designed to go beyond the stand-alone BI tools of the past by handling a broad range of data modeling and analysis tasks. The company's software is also finding its way into a growing number of embedded applications.

Data scientists' chief task is developing analytical and predictive data models that underpin business analysis systems. But before they can build those models, huge volumes of data often must be collected, curated and cleaned – a time-consuming job.

"If you talk to any data scientist, they will tell you this is where they spend a lot of their time," said Daniel Mintz, Looker's own data scientist, in an interview with CRN.

Looker first debuted Action Hub in Looker 5 last September. The new tools and integrations in Action Hub announced this week are designed to improve data science workflows, accessing disparate datasets and helping prepare data for building and testing predictive analytical models.

The new capabilities include a software developer kit for the R programming language and connections for the Python programming language. The software can combine data from multiple sources into a single analytic view, query and stream data from massive data sets for data modeling, and perform advanced statistical analysis directly within Looker.

A key enhancement is Looker's new integration with TensorFlow, an open-source software library and machine-learning framework for dataflow programming and high-performance numerical computation. It's also integrated with Big Squid's Kracken software that automates machine learning for data analytics.

Looker has also partnered with a number of data science software developers, including Data Robot, to improve linkages between their tools and the Looker platform.

Currently partners in the Looker Partner Network, including solution providers, systems integrators, consulting service companies and OEM technology partners, largely focus on working with business analytics managers and line-of-business managers. Mintz said partners would benefit from the new capabilities by expanding their potential customer audience to data scientists within customer businesses and organizations.