Tableau is introducing a new artificial intelligence tool that promises to make complex data
analysis and visualization tools more accessible to a wider range of customers.
Tableau, headquartered in Seattle, is a long-standing analytics company whose platform seeks to enable both data specialists and self-service business users to explore and analyze data to influence choices.
Salesforce, which paid $15.7 billion for Seattle-based Tableau in 2019, is integrating Einstein Copilot and other AI tools throughout its platforms, including its core customer-relationship management offerings.
Tableau originally announced ambitions to develop generative AI capabilities at its annual user conference in May 2023, when it unveiled Tableau Pulse and GPT.
Einstein Copilot for Tableau is designed primarily for data analysis and goes beyond simply asking natural language questions about data. The new copilot incorporates several tools to assist business users and data analysts in getting past the dreaded “blank page” condition, in which they are unsure what to do next or how to conduct a certain activity. Einstein Copilot for Tableau includes recommended questions to assist users figure out what to ask about a certain data source. There is also a conversational data exploration feature to let people go further into the results.
“We don’t want to come out and say to the user – you need to write better prompts,” Tableau’s chief product officer, Southard Jones, told VentureBeat. “So we spent a lot of energy and time making sure that when an analyst is in the traditional Tableau pills and shelves experience and they ask questions that they can get responses back or we guide them to very specific answers.”
In Tableau, a pill refers to the data kinds that a user enters into a workflow, whereas shelves relate to the columns and rows of data being evaluated.
However, like Einstein Copilot for Tableau, the majority of the AI helpers that have been deployed are still in the preview or testing stage, according to Doug Henschen, an analyst at Constellation Research. Even those that are widely available are so new that it is impossible to determine how one compares to another.
Meanwhile, many analytics customers are hesitant to adopt new tools because to worries about language model security and the expense of developing and operating generative AI systems, according to Henschen.