Salesforce is adding a bit of agentic magic to Tableau. The company announced it’s bringing agentic analytics to the data visualization software through a new offering called Tableau Next. It’s all about helping executives get faster, sharper intelligence to drive decisions. The tool leverages AI to analyze a company’s data better and provide actionable insights through existing systems while featuring open and composable architecture suitable for the enterprise.
What is agentic analytics? Salesforce defines it as the ability of an agent or software to examine data and allow humans to engage and uncover answers. However, the responses won’t be rules-based; instead, they will be generated using reasoning and can accept natural language prompting.
Closing the Data Insight Gap
The problem Tableau Next is designed to solve is helping business leaders feel more confident in their organization’s data. A recent study by the company of 500 business leaders found that most respondents felt their jobs depended on their data literacy. “Many of them are not confident in their data themselves and only half of them are sure they can even find or interpret the data that they need,” Southard Jones, Tableau’s chief product officer, remarked during a press briefing last week.
“There is a gap between ‘I need to become seen as a data literate person. I need data within 30 minutes to deliver accurate business decisions. Yet, only half of us have access to analysts to give me that data, and only half of us actually feel they can interpret that data on their own.’ So there’s an opportunity for technology to fill in that gap, and that gap isn’t one just based on a need to access data,” he added.
In other words, executives may have the data, but turning raw numbers into actionable insights is still slow. Without real-time interpretation, organizations often lose confidence in what the data tells them. Tableau bets adding AI agents can be the solution these business leaders seek.
The Tableau Next Platform
Jones calls this offering an “evolution” of the existing Tableau platform. All the visualizations on there before are carried over to Tableau Next. “This is not a separate platform,” he emphasizes. “This is an extension that allows you to leverage your existing work in Tableau.”
Tableau Next, first previewed in March, has four layers, starting with an entirely open platform featuring all of an organization’s data. Jones states that Tableau previously only analyzed structured data, but now it can process unstructured data and streaming data (e.g., social media content). “In the past, that hasn’t really been available—it was only for a few people that had access to that. Business leaders should have access to that. So, we’ve broadened that spectrum to just about any type of data available inside…and outside of your company.”
The next layer is what Tableau calls the AI semantic layer. It’s the part of Tableau Next that converts the raw data and business conversations taking place inside an organization into actionable insights. Jones equates it to a restaurant menu in which, in the kitchen, there are raw ingredients with different names that we might not understand. A chef will use them to create a dish that we recognize. “There’s a ton of ingredients in it, but we don’t know there’s all those ingredients,” he shared. “We just know that the dish tastes good, and it’s the language we speak. That is what a semantic layer is for all our businesses.”
On top of that sits the visualization layer, which Tableau has traditionally been known for. The company has enhanced it by making visualizations “composable and modular so that you can place them wherever someone works.” According to Jones, every visual built inside Tableau Next is designed to be placed elsewhere, from an application to a collaboration platform like Slack, email, or your phone. Business intelligence “shouldn’t be a destination. It should just be where you work, a little bit like your car navigation system. You don’t need to go somewhere else to read a map. It is read to you. It is told to you. That’s the visualization layer within Tableau Next.”
The top layer of this platform is a workflow engine that generates insights tied to actions. Examples Jones provided include approving campaign spending increases or where you’re collaborating with a teammate to provide feedback. Regardless of the use case, he pointed out that the insights will be directly tied to an action in a workflow.
The Agents of Tableau Next
To help organizations use this new offering, Tableau is releasing three new AI agents: Data Pro, Concierge, and Inspector.
Data Pro acts like a virtual analyst, examining the company’s raw data and translating it into the business language you speak. Jones disputes that Tableau is looking to displace the workers who help with calculations and the ETL (extract, transform, lead) process. Instead, the company claims its agent offers a way to expedite accessing data faster and deliver insights faster.
Concierge is like a hotel concierge but for an organization’s data. You can ask it questions about the data and it will return immediate and reliable answers.
Lastly, the Inspector agent functions as proactive data monitoring and insights. Jacob believed this bot has redefined how companies use BI: Tools have long assumed humans know what questions to ask. But if the question isn’t asked, the business misses the insights. To overcome this, the Inspector examines all the data and runs an algorithm against it to identify patterns and surface them for the appropriate people at the right time.
These three agents were built on top of the Agentforce platform. Salesforce has integrated the AI service into Tableau to bring reasoning to help answer data questions.
“Tableau Next helps businesses achieve faster, more impactful results with their data,” Tableau Chief Executive Ryan Aytay said in a statement. “We’re shifting from basic reports to a world where AI is a collaborative decision-making partner. By combining AI agents with trusted data and easy-to-use tools, we’re making data accessible to everyone and transforming the data-to-action process into an automated, proactive, and insight-driven cycle.”
The On-Demand, AI-Powered Data Analyst for Businesses
“The biggest challenge for us is getting data at the right time, at the right moment, it being accurate, having high confidence in the fidelity of it, and being efficient in how we get that data,” Ravi Malick, Box’s global chief information officer, remarked. He points out that for companies that have data silos, there’s a battle to consolidate the information and move them into a place where “we can…effectively use it, make it available to business leaders and business users and…then have the right layer on top that provides the interaction with that data.”
Malick points out a common frustration at companies like Box: an executive asks a question and hears, “I’ll get back to you”—often followed by a vague timeline of hours, days, or even longer. Tableau Next promises to be an on-demand AI-powered analyst that can give business leaders the real-time insights they want to take action promptly.
Jones insists Tableau Next won’t replace human analysts—in fact, he believes it will increase demand for them. Their role will evolve, with them spending less time “preparing data, cleansing data, things that aren’t really value-added tasks, and they will spend more time engaging with the business partners and helping [them] make decisions at the right time.”
He predicts that analysts’ scope will change, just like software development nearly two decades ago. Jones highlights that back then, software developers were divided into front-end and back-end. But today, everybody is a full-stack developer, thanks to tools being introduced to broaden their scope and help increase productivity. When it comes to analysts, tools like Tableau Next will have similar benefits and lead them to invest time working on “the things that are closer to the right side of the equation—that value-added side of the equation—and reduce some of that left side of the equation—the data prep side.”
Tableau Next is now generally available. However, the three agents—Data Pro, Concierge, and Inspector—will be made public later this year.
Featured Image: Tableau Next. Credit: Salesforce
Leave a Reply
You must be logged in to post a comment.