Salesforce Tableau looks beyond business intelligence dashboards with AI-powered Pulse

Business intelligence (BI) has long been heavily reliant on dashboards to help organizations make sense of data, but there’s a lot more meaning and value from data that can be uncovered with the use of artificial intelligence (AI).

Today Salesforce announced the general availability of its Tableau Pulse technology, bringing the power of multiple forms of AI, including generative AI to help organizations get more business intelligence insights from their data. Tableau Pulse was first announced in an early preview in May 2023, with a limited beta in December 2023. Tableau Pulse uses natural language queries and data visualizations to surface important metrics, trends and recommendations specific to each organization’s data. 

With the general availability release, Tableau is also expanding Pulse with enhanced capabilities including new insights features, a news digest of data insights for Slack and a metrics layer to help organizations define and understand what the data means.

“We spent an inordinate amount of time with customers and other users, giving them mock-ups and going through tests to see what is working and what’s not working and figuring out how to engage people that aren’t necessarily data dorks,” Southard Jones, Chief Product Officer, at Tableau told VentureBeat.

How Tableau Pulse uses AI to find insights that dashboards don’t show

The core Tableau platform has long been collecting business data that data scientists and business users can use for business intelligence. The challenge that Pulse is aiming to solve is how to surface the right insights.

“The concept behind Pulse is essentially giving people metrics that they live and die by, or are hired and fired on, that they manage their business with,” Jones said. “It’s about letting them know when things change in a way that they can take action quickly.”

Jones noted that in theory any of the insights that Pulse surfaces could also be in a dashboard. The difference is that the Pulse makes it easier for a user to query the data and also works more proactively. According to Jones, a common issue with BI platforms overall is that the dashboards show problems that have already occurred. In contrast, the goal with Pulse is to be able to surface insights and issues before they become problems.

Jones emphasized that rather than relying on creating dashboards, what Pulse aims to do is help business users and data analysts to more easily expose things that are happening in the business through data.

During the preview period for Pulse, Jones said that one early user told him that Pulse helped to identify a product return issue promptly that managed to save the company money.

“There was a particular item that was getting returned frequently,” Jones said. “But they did some more research and they found out the sizing was a little bit off and they were able to change that in time for the holiday season.”

New metrics and insights digest add value 

The new Tableau Pulse Slack Digest feature is also a key part of looking beyond the traditional dashboard approach to business intelligence.

Like Tableau, Slack is a Salesforce company. The Slack digest is a news feed-like experience delivered through Slack. It provides headlines and summaries of insights in a digest format to engage users who may not access dashboards regularly.

Sitting underneath the insights are the metrics that the organization wants to measure. A common challenge with data analytics is when organizations don’t have clear definitions for specific metrics which can lead to confusion. 

The metrics layer creates a single definition for each metric across all of an organization’s Tableau workbooks. This provides governance and consistency over how metrics are tracked. It also allows natural language queries about metrics to be better understood through context and routed to the correct metric definition.

Pulse isn’t just about Generative AI

The idea of having a generative AI assistant or copilot to query data is increasingly commonplace across enterprise applications.

Jones emphasized that Tableau Pulse isn’t just an assistant that is bolted onto the platform, rather he said it’s deeply integrated with the overall Tableau platform. He also noted that Tableau is using traditional AI that it has developed for insight summaries based on statistics and models. Part of that approach comes from Tableau’s acquisition of Narrative Science back in 2021, to help build out natural language summaries of data.

Tableau Pulse also makes use of Salesforce’s Einstein GPT as a gen AI layer for content creation. Jones explained that  Salesforce integrates a trust layer for security and to check for any potential hallucination risk.

“The problem we’re trying to solve is that there’s a lot of users in companies today who are not leveraging data, that could be using data to make better decisions,” Jones said.

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.