Tableau enters the agentic AI era with the launch of Next
The new platform features agents addressing needs such as data preparation and natural language-based analysis, and includes a semantic layer that targets relevancy and accuracy.
With the launch of Tableau Next on Tuesday, the AI generation of Tableau has arrived.
First unveiled in preview in September as Tableau Einstein, Tableau Next is an AI-based analytics platform that combines existing capabilities from the vendor and its parent company, Salesforce, with new features including generative AI-powered agents.
The platform, available as part of the Tableau+ subscription, aims to automate repetitive tasks, proactively deliver insights and enable users to take action within the context of workflows. It features an open data layer that includes unstructured and streaming data, an AI-powered semantic layer, AI-first composable visualizations, and an action layer that integrates with external applications to enable users to trigger actions.
In addition, Tableau Next includes three agents -- Data Pro, Concierge and Inspector -- currently in preview that will assist users and take on tasks to enable faster and more comprehensive analysis.
With the analytics sector transitioning to agentic AI, Tableau Next shows that Tableau is making the change from a traditional BI platform to AI, according to David Menninger, an analyst at ISG Software Research. As a result, it's a significant new platform.
"Tableau Next represents the reengineering of Tableau," he said. "Agentic workflows [decompose] Tableau data and analytics tasks into individual components that can be assembled into a variety of workflows. The ultimate goal is to automate more of the analytics process."
Tableau Next represents the reengineering of Tableau. Agentic workflows [decompose] Tableau data and analytics tasks into individual components that can be assembled into a variety of workflows. The ultimate goal is to automate more of the analytics process.
David MenningerAnalyst, ISG Software Research
In addition to Tableau, Google Cloud with Looker is another vendor that has made agentic AI central to its platform as generative AI has quickly evolved since OpenAI's November 2022 launch of ChatGPT represented a significant improvement in the technology.
A new Tableau
While enterprises are increasing their investments in generative AI capabilities to make employees better informed and more productive, data management and analytics vendors are similarly developing generative AI tools to make their complex platforms easier to use and more efficient.
At first, those generative AI tools were bots that enabled users to query and analyze data using natural language and do time-consuming tasks such as generate code and document work. Now, agents -- which, unlike bots that act only when prompted, can act autonomously to surface insights and execute processes -- are the vanguard.
Data Pro, available in June, is a data preparation agent that provides suggestions for complex steps such as data transformation and automatically takes on some of the work on its own. Concierge, also available in June, is a natural language query and response agent that includes recommended actions. Inspector, generally available by year-end 2025, automatically monitors data for changes, analyzes trends and makes predictions so that users can take proactive steps.
Each is underpinned by Tableau Semantics, a semantic layer that provides Tableau Next's agents with their understanding of business data to generate relevant, accurate outputs.
Rebecca Wettemann, CEO of research firm Valoir, noted that Tableau has long attempted to simplify analysis of complex data. Tableau Next is the evolution of that approach, simplifying analysis and enabling users to take action based on their analysis.
"Next is about leveraging more autonomous AI capabilities to do more of the heavy lifting like data prep and recommendations," she said. "The other key piece is the workflow engine that enables users to kick off automated actions based on a particular insight, which is key for moving from just insights to intelligent actions."
Tableau's research shows that agentic AI has the potential to improve the relevance, reliability and timeliness of data, according to Southard Jones, Tableau's chief product officer.
"The three agents were picked based on customer interviews and the research we've done," he said during a virtual press conference on April 11. "In the product, we talk about 'jobs to be done,' and we want to help [users] do those tasks more productively in a more scalable and higher-quality way."
Tableau Next features generative AI agents such as Concierge that enable users to query data using natural language and receive visualizations and summaries in response.
Ravi Malick, senior vice president and global CIO at Box, a Tableau customer, said he feels pressured to act quickly based on data.
"In some cases, you can't wait an hour," he said during the virtual press conference.
Malick has been using Tableau Pulse, a generative AI-powered insight generator launched in February 2024, and experimented with Tableau Next while it was in preview. Now, he said he's planning more widespread use and highlighted the benefits of Concierge.
"I immediately thought about how many questions I ask my team that are data-related, and the ability to [ask Concierge] is incredibly exciting," Malick said.
Menninger called out Concierge and Data Pro as agents addressing real user needs, while Wettemann noted the potential value of Inspector.
Looking ahead
Tableau's next agent could be one that helps data stewards develop their semantic layer, according to Jones.
Wettemann, however, suggested that with generative AI capabilities becoming common among analytics vendors, vendors such as Tableau could provide more thought leadership rather than add more tools. Customers could use help understanding how they can benefit from generative AI without having the technology take over their jobs.
"I expect Tableau will deliver more agentic skills," Wettemann said. "But even more important is going to be the best practices and communication around how Tableau champions can use these capabilities to ... show the value of what they're delivering when much of their day-to-day work has the potential for automation."
Menninger, meanwhile, said that while Tableau is now among those analytics vendors at the forefront in terms of delivering generative AI capabilities, challenges remain. As organizations turn certain tasks over to agents, more needs to be done to enable agents to work collaboratively.
"One of the next big challenges around generative and agentic AI is multi-agent processing," Menninger said. "As an industry, we need to develop agent-to-agent protocols. ... Supporting [protocols] or other industry standards will be critical to the evolution of agentic AI."
Eric Avidon is a senior news writer for Informa TechTarget and a journalist with more than 25 years of experience. He covers analytics and data management.
