New Data Science Workflows in Looker


Santa Cruz, CA – May 15, 2018 – Looker, a leading data platform company, today announced capabilities that improve and optimize data science workflows. Looker turbocharges the data science stack and gives data scientists greater freedom to work on high-value tasks.

Looker already delivers clean data, at scale, for data scientists to input into their models and then present the insights in understandable and actionable dashboards and reports. Now, Looker has further optimized this workflow with an SDK for R and connections for Python, as well as, streamed and merged results, Google TensorFlow integrations, and clean, visual recommendations for users. Looker also continues to partner with best-of-breed data science vendors to ensure seamless integration.

“Cleaning and wrangling data is a waste of time for data scientists,” said Frank Bien, CEO of Looker. “Looker makes data science workflows vastly more efficient, so data scientists and architects can spend more time on high-value model building, create greater business impact, and move on to the next problem faster.”

In addition to Looker’s ability to deliver clean data for data scientists and presenting insights to users across an organization, new capabilities include:

  • R SDK  – Easily leverage data from Looker while working with R and RStudio
  • Python connections – Easily leverage data from Looker while working with Python and Jupyter Notebooks
  • Machine Learning/Artificial Intelligence Partners – Integrate with best-of-breed partners to make the Data Science workflow more efficient, including Big Squid and TensorFlow from Google
  • Stream results – Send the entire results of your query for use in data science modeling
  • Merge results – Combine data from multiple different sources in the Looker frontend
  • Statistical functions – Calculate advanced statistics directly in Looker
  • Suggested analytics – Get intelligent recommendations from the Looker home page and autocomplete

“We use Looker as a single source of truth for clean data about our clients, and rely on it when building predictive models or collaborating on metrics with other teams internally,” said Julia Silge, data scientist at Stack Overflow. “It’s a massive time saver for us, reducing steps that used to take hours to only a few minutes.”

“DataRobot and Looker have complementary approaches to data: we simplify, model, and automate user workflows for big data analytics and machine learning to improve efficiencies,” said Seann Gardiner, EVP, Business Development at DataRobot. “Our joint customers see dramatic increases in the efficiency of their machine learning workflows, more than 500% in some cases. With tools like DataRobot for deploying AI and Looker for curating data, users can build powerful, enterprise-grade, automated machine learning more quickly than ever.”