Spotfire unveils generative AI plans, adds to pre-built apps
The analytics specialist introduced the private preview of Copilot, which will enable customers to ask questions of their data and receive responses without having to write code.
Spotfire unveiled new generative AI capabilities, including its version of a copilot that enables customers to interact with data using natural language rather than code.
In addition, Spotfire -- a business unit of Cloud Software Group (CSG) that specializes in analytics -- revealed new Spotfire Data Functions, Mods and Apps aimed at helping users more easily reach insights and take action based on those insights.
Spotfire introduced the private preview of Copilot and its generally available additions to Data Functions, Mods and Apps on Sept. 19 during its Analytics Forum, a virtual user conference.
Spotfire had long been part of Tibco, which is a subsidiary of Vista Equity Partners. In 2022, Vista acquired Citrix and subsequently merged Citrix with Tibco. Following that merger, Vista reorganized Tibco, which included forming CSG in late 2022.
Tibco is now a business unit of CSG focused on data integration and data virtualization. Meanwhile, companies that Tibco acquired over the years to add and enhance analytics capabilities, including Spotfire in 2007 and IBI in 2020, are now also their own business units within CSG.
New capabilities
Generative AI has been the dominant trend in analytics for the better part of the last year, ever since AI vendor OpenAI launched ChatGPT in November 2022, marking a substantial leap in generative AI and large language model capabilities.
Since then, many data management and analytics platform updates have focused on integrations with generative AI platforms and adding generative AI capabilities.
Sisense was among the first, unveiling an integration with ChatGPT in January 2023. Most recently, Oracle introduced generative AI capabilities on Sept. 20 -- the day after Spotfire unveiled Copilot. Also recently, Qlik revealed a new suite for generative AI and machine learning in early September, and Domo launched a new suite for generative AI development in late August.
Meanwhile, ThoughtSpot, Tableau, Microsoft, Google, Databricks, Snowflake and more have all made generative AI the focal point of their product development.
The reason is generative AI's potential to make data management and analysis easier and more widespread.
For decades, data and analytics growth has been stalled. Depending on the study, it's estimated that only a quarter to a third of employees within most organizations use data as a regular part of their job.
One hindrance of growth is that most data management and analytics tools are designed for data experts, requiring users to write code and making the tools difficult to learn. Another impediment is that data analysis has required data literacy training, and many organizations have not invested in programs that teach employees how to interact with data.
Spotfire's introduction of Copilot, therefore, is a significant move for Spotfire customers, according to Kevin Petrie, an analyst at Eckerson Group.
Using Copilot, customers will essentially be able to converse with their data. They will be able to ask questions, receive data visualizations and summaries in response, and ask follow-up questions to further analyze data.
Copilot, meanwhile, is able to understand the intent of a user's question and then use curated domain-specific data rather than public data to deliver its response -- an approach that reduces the frequency of incorrect responses that plague models trained on public data, according to Petrie.
"Tibco and now Spotfire worked on language models before ChatGPT debuted last fall," he said. "So this team has longer experience with natural language processing than some other vendors. Spotfire also is taking the right approach by aligning with what we call retrieval-augmented generation ... that helps ground language models in reality and reduce the likelihood of hallucinations."
Heterogeneous streaming and virtualization support in particular differentiates Spotfire's BI and Copilot offerings from competitors.
Kevin PetrieAnalyst, Eckerson Group
Spotfire is not the first platform to use the term copilot to describe its natural language query and analysis capabilities. Microsoft is implementing Copilots throughout its array of products, including Power BI.
Whether called Copilot or another name, vendors including Google and AWS have also introduced generative AI query and analysis capabilities for their BI platforms. But that doesn't make Spotfire Copilot any less significant for Spotfire customers, Petrie said.
In fact, he noted that Spotfire's Copilot stands out from other vendors' generative AI tools because of its tie-ins to Spotfire's data streaming and data virtualization capabilities.
"Heterogeneous streaming and virtualization support in particular differentiates Spotfire's BI and Copilot offerings from competitors such as Microsoft," Petrie said.
As Spotfire develops Copilot, one of the decisions the vendor needs to make is whether to make it a feature within the platform or an open source community download, according to Michael O'Connell, Spotfire's chief analytics officer.
However it's ultimately delivered to customers, Copilot is intended to better enable analysis that leads to action. So are the additions to Spotfire Mods, Data Functions and Apps, O'Connell continued.
Spotfire Mods, first unveiled in 2020, are pre-built applications that enable customers to build analytics products such as dashboards and visualizations without requiring them to write code. Data Functions, revealed in 2021, are an extension of Mods designed to simplify development of predictive models and other AI and machine learning assets.
Spotfire Apps are visualization templates that customers can use to save time developing charts and dashboards. Previously known as Tibco Accelerators, they now have a new name following the formation of CSG and Spotfire becoming its own business unit.
"The big theme is enabling customers to self-serve assets and functionality that allows them to rapidly identify insights and take actions to grow their business," O'Connell said.
Next steps
With Copilot now in private preview and Spotfire continuing to add pre-built templates in the forms of Apps, Data Functions and Mods, ease of use will continue to be a goal for Spotfire, according to O'Connell.
Tibco historically catered to data engineers and other experts rather than business analysts. That began to change a bit over five years ago when the vendor added more self-service capabilities to Spotfire.
Since then, Spotfire has been updated to include pre-built applications such as Mods and Data Functions that make development and analysis simpler. Most recently, more easily enabling actions became a goal, and Spotfire 12 included capabilities that connect Spotfire to other work applications and allow users to trigger actions within those other applications directly from Spotfire.
"It's about making our product easy to use to do sophisticated things," O'Connell said. "Our stuff is not just backward-looking reports. It's about real-time data, embedded data science, and highly functional visual analytics for finding insights and taking actions."
Petrie, meanwhile, said he's curious to see how Spotfire continues to integrate AI and machine learning with business intelligence.
The platform includes a mix of capabilities that target not only data scientists and data engineers, but also business analysts, he noted. How Spotfire enables those different personas to work together will be worth watching.
"Spotfire's portfolio puts it in a good position to help data analysts, data scientists and machine learning engineers collaborate to build converged outputs and workflows," Petrie said.
Eric Avidon is a senior news writer for TechTarget Editorial and a journalist with more than 25 years of experience. He covers analytics and data management.
