Spotfire has put its built-in AI at the center of its latest quarterly update, previewing Spotfire 15.1 with a natural-language assistant, reusable analytical agents, and the ability to reach into other operational systems from inside an analysis. The company, a business unit of Cloud Software Group, framed the release around a problem specific to industrial work: generating an analysis is easy, but trusting one is not, and in a plant an unverifiable answer can mean scrapped lots or rejected batches.
The headline addition is Ask Spotfire, arriving with 15.1, which lets an engineer describe what they need in plain language and have Spotfire build or modify the analysis. Around it sit Insight Agents and Skills, which let a team capture an analytical method once so anyone can rerun it on new data, including agents built specifically for manufacturing and energy. The stated aim is to keep the domain expert in control: teams ask in their own words, check the result against their real data, and reproduce the method, rather than taking an AI answer on faith.
For process and quality engineers chasing yield, the semiconductor-specific work is the concrete part. Spotfire 15.1 adds wafer layout detection, statistics layers, and exposure-field visualization, so an engineer can analyze spatial wafer data and work toward root cause in one visual environment instead of moving between tools. "Spotfire brings together the fragmented, spatial, and temporal data of industrial operations with validated methods and visible evidence, so experts can decide whether an answer is true," said Niklas Amberntsson, director of product management at Spotfire.
The release also adds support for Model Context Protocol, the open standard for connecting AI to other software. In Spotfire that means linking its AI to systems such as maintenance planning or ticketing and triggering an action from within the analysis, rather than exporting a finding and acting on it elsewhere. The same standard has been showing up across industrial software this autumn as vendors wire AI agents into existing tools.
Governance is the throughline. Organizations choose which AI models Spotfire uses and which groups can use them, with usage logged and auditable. "Producing an analysis is no longer the hard part; trusting it is," said Tobias Lehtipalo, VP of product management at Spotfire, framing the update as letting engineers work under the governance their organization already runs. The quarterly update builds on capabilities shipped earlier in the year, including the Push Compute engine for running large-scale analytics inside cloud data platforms, which Spotfire launched on Snowflake in August.
