ThoughtSpot has expanded its integrations with Snowflake Cortex AI and Snowflake Semantic Views, extending how its analytics agents work inside Snowflake.
The changes tie ThoughtSpot's Spotter product and related agents more closely to Snowflake's AI and semantic tools, aiming to let customers use governed business logic across both human users and software agents. They are also designed to keep analytics activity within Snowflake's security boundary while drawing on data and definitions already managed there.
At the centre of the update is Spotter, ThoughtSpot's analytics agent, which now integrates with Cortex Analyst and Cortex Agents. Customers can access and use insights generated through Snowflake's Cortex tools from within ThoughtSpot's interface, rather than moving between separate systems.
ThoughtSpot is also adding support for native visualisations linked to Cortex-generated answers. Users will be able to turn text-based responses into interactive charts and Liveboards inside ThoughtSpot, extending Snowflake's AI outputs into more conventional business intelligence formats.
Another part of the update is a bring-your-own-model approach based on Snowflake-hosted large language models. Customers will be able to use the platform with Snowflake LLMs already approved for their environments, while keeping sensitive data inside Snowflake rather than moving it to external systems.
Semantic layer
A second part of the announcement focuses on data modelling and semantic management. ThoughtSpot now provides native support for Snowflake Semantic Views, allowing business logic, metrics, and relationships defined in Snowflake to flow directly into ThoughtSpot.
It is also introducing bi-directional semantic management through an integration with Snowflake CoCo. Customers can import semantic definitions from Snowflake into ThoughtSpot and export ThoughtSpot models back into Snowflake, including additional AI context and memory created inside ThoughtSpot.
That shared semantic layer matters because many companies are trying to deploy AI systems across multiple data sources without losing control of the definitions behind core metrics and business terms. In practice, mismatched definitions can lead to inconsistent outputs across dashboards, reports, and AI-generated answers.
The expanded integration is intended to reduce that risk by making Snowflake the common source for semantic definitions while allowing ThoughtSpot's products to work from the same governed structure. It also covers structured and unstructured data.
Agent suite
The update applies across ThoughtSpot's broader set of software agents. These include Spotter for enterprise analytics, SpotterModel for data modelling, SpotterViz for dashboard assembly, and SpotterCode for software development work linked to embedding ThoughtSpot functions into other applications.
SpotterModel is aimed at analytics engineers and data analysts responsible for the semantic layer. It uses natural language prompts to build and edit reusable data models, which could reduce manual modelling work for data teams.
SpotterViz is designed to automate the creation of Liveboards, including layout, organisation, styling, and publishing. SpotterCode, meanwhile, is intended to help developers embed ThoughtSpot functions into custom applications with AI-assisted code generation in development environments such as Cursor, Claude Code, and VS Code.
These products can now be used more uniformly inside the Snowflake ecosystem because they are grounded in Spotter Semantics and integrated with Snowflake Semantic Views. ThoughtSpot also supports Snowflake Interactive Analytics, used for high-concurrency and real-time analytics workloads.
ThoughtSpot's platform and agent products are available through the Snowflake Marketplace, where customers can deploy them using Snowflake credits. The route reflects a broader pattern in enterprise software, as analytics and AI vendors seek tighter commercial and technical alignment with large cloud data platforms rather than asking customers to run separate stacks.
Francois Lopitaux, Senior Vice President of Product Management at ThoughtSpot, outlined the company's view of how the market is changing.
"Businesses have moved beyond wanting top line insights driven by general AI agents. The winners in the enterprise AI era will be the organisations whose agentic systems are grounded in governed business context across structured and unstructured data," said Lopitaux.
"By further integrating Spotter with Snowflake Cortex AI and allowing customers to bring their own fine-tuned Snowflake LLMs, we are giving enterprises the accuracy, control, and data sovereignty they need to operationalise trusted AI at scale," said Lopitaux.
Snowflake also framed the tie-up around data governance and consistency in AI outputs.
"The combination of Snowflake Semantic Views and ThoughtSpot's Spotter Semantics represents a massive leap forward for data governance and analytics. Organisations can now define their semantic context for AI and BI once in Snowflake and use ThoughtSpot as a trusted intelligent context layer for AI and agents to scale those definitions across the enterprise. This helps ensure that agentic workflows and AI-generated insights are based on a single, governed source of truth," said Josh Klahr, Director of Product Management, Analytics at Snowflake.