ThoughtSpot Inc. today launched a new version of its Analyst Studio data analytics platform, introducing capabilities intended to help organizations prepare data for artificial intelligence workloads at lower cost while maintaining governance controls.

A new feature called SpotCache provides a spreadsheet-style interface for data preparation and an agent-based data preparation tool that uses natural language interactions to assist analysts and data teams. ThoughtSpot is positioning the release of the platform (pictured) as a response to what it describes as an “AI readiness gap,” in which organizations are adopting AI agents and natural language interfaces faster than they can prepare the underlying data to support them.

“The journey to effective AI starts with data readiness, but for too many teams, that journey is stalled by rigid tools and unpredictable cloud costs,” said Anjali Kumari, vice president of product management at ThoughtSpot.

Avoiding the query trap

SpotCache is a caching layer designed to reduce the cost impact of repeated AI-driven queries against cloud data warehouses. It addresses AI workloads where user questions may result in multiple queries.

It maintains prepared datasets that can be queried without incurring additional data warehouse consumption charges. “It’s a layer that persists your data after you have prepped it to ensure that you can query it without worrying about increasing your compute cost,” Kumari said.

Data from the source database is loaded into DuckDB, an open-source column-oriented relational database management system designed for high performance, running in ThoughtSpot’s cloud. Users can run as many queries as they want, but the dataset size is limited to specified tiers.

It’s up to customers to decide which datasets are the best candidates for caching based on where cloud costs are highest.

SpotCache also supports flexible refresh scheduling that can range from hourly to weekly. It doesn’t yet support automatic updates triggered by upstream data changes but Kumari said such a feature may be added later.

Security controls

Administrators can apply role-based access controls, row-level security and column-level security to cached data with full audit trail capabilities. SpotCache doesn’t currently inherit security controls directly from the cloud data warehouse, meaning they must be applied manually.

ThoughtSpot also announced a new spreadsheet-style interface intended to make data preparation more accessible to business analysts. Kumari said the interface is Excel-like and supports ThoughtSpot-specific extensions like user cohorting and complex formulas.

Also new is an agentic AI capability that assists users in integrating data preparation more closely with the analytics consumption layer so that data issues can be identified and corrected in context.

“Data prep is becoming more and more of a bottleneck,” Kumari said. “We’re looking to make it continuous.”

The company said Analyst Studio, including SpotCache, is now generally available for ThoughtSpot Analytics and ThoughtSpot Embedded customers, with additional capabilities expected to roll out later this year.

Image: ThoughtSpot