Petya Petrova - Fotolia
The top feature of the ThoughtSpot 5 update, which was unveiled at ThoughtSpot's 2018 user conference this week, SearchIQ builds on the platform's relational search functionality to add voice and conversational analytics capabilities.
The relational search allows users to "ask precise questions and get precise answers," said Amit Prakash, ThoughtSpot co-founder and CTO, in an interview.
With SearchIQ, which is currently in public beta, users can "ask questions in more conversational ways," while still getting those precise answers, he continued.
Ask a question, get an answer
Founded in 2012 and based in Palo Alto, Calif., the analytics software vendor has long included search functions and basic natural language processing (NLP) technology in its BI platform. With ThoughtSpot SearchIQ, users now are able to search using natural language or voice on desktop, mobile or connected devices.
Features sit on top of the platform, Prakash explained, in a way that takes conversational language and turns it into language the relational search technology can understand.
"It's about supporting true [verbal] conversations with your data," said Wayne Eckerson, an analyst at Eckerson Group, in an email from ThoughtSpot's Beyond 2018 conference.
Eckerson said while he's not sure how the product will be adopted, "most likely it might gain traction with folks who need hands-off access to data: people on shop floors, mobile executives, field technicians and sales people."
Users can go in and change queries to customize the results, Prakash said. Machine learning algorithms will learn from these tweaks, giving users personalized results and enabling higher search intent percentages after a few days of use, he added.
The easy-to-use -- according to ThoughtSpot -- tool is meant to give casual users and citizen data scientists more analytic power.
"If you want to ask a simple question and you're a casual user, it works," Prakash said. If you're an advanced user and want to use advanced functions, "it works, too."
A possible trend
ThoughtSpot SearchIQ arrives on a wave of analogous NLP products from similar vendors. Tableau, for instance, recently released its Ask Data tool, and IBM added an AI assistant to its Cognos Analytics platform.
Wayne Eckersonanalyst, Eckerson Group
However, Prakash claimed that ThoughtSpot differs because "we have been building it from the ground up -- it's a product that is built for search."
What that means is that ThoughtSpot 5 and ThoughtSpot SearchIQ enable, for example, users to input billions, versus millions, of rows of data, and get fairly quick responses to queries, the company claims.
When a user puts data into ThoughtSpot, there could be many rows, Prakash said. The software indexes all that data and even if the user mistypes something, the system can generally figure it out, he asserted.
With that in mind, users are also able to ask ThoughtSpot SearchIQ complex questions with multiple variables, which will return precise answers, according to the company.
More with ThoughtSpot 5
Apart from SearchIQ, ThoughtSpot 5 adds more instant insights and recommendations from its SpotIQ tool, an AI-powered component the vendor added last year, as well as better R integration and enhanced visualizations, the company said.
Also unveiled at the ThoughtSpot conference, held Nov. 13 to 15 in National Harbor, Md., was news that ThoughtSpot has been certified to run its Falcon engine on Google Cloud Platform, and has enabled more integrations with the Google Cloud Machine Learning Engine.
ThoughtSpot "wants to establish partnerships with many vendors -- they want to be a platform that integrates with other tools," Eckerson said. "For Google, it's a good way to access data in the cloud."
"Most data will go to a cloud platform in the next 10 years -- so ThoughtSpot needs good connectors to data anywhere," he noted.