ThoughtSpot expands its integration with Snowflake Cortex AI and Snowflake Semantic Views to power Spotter analytics agents inside Snowflake, enabling governed business logic across both humans and software agents.

BYOM support lets Snowflake-hosted LLMs work with sensitive data staying inside Snowflake, avoiding data movement to external systems.

A shared semantic layer ensures consistent definitions across dashboards, reports, and AI-generated outputs when multiple data sources are in play.

SpotterModel targets analytics engineers with natural language prompts to build and edit reusable semantic data models.

Native visualisations linked to Cortex-generated answers let users turn text responses into interactive charts and Liveboards within ThoughtSpot.

SpotterViz automates Liveboard creation—from layout to publishing—while SpotterCode helps developers embed ThoughtSpot functions and AI-assisted code in environments like Cursor, Claude Code, and VS Code.

Bi-directional semantic management via Snowflake CoCo enables importing semantic definitions into ThoughtSpot and exporting ThoughtSpot models back to Snowflake, including AI context and memory from ThoughtSpot.

Agent suite updates unify Spotter, SpotterModel, SpotterViz, and SpotterCode for use within Snowflake and reference Spotter Semantics.

The integration leverages Snowflake Interactive Analytics for high-concurrency and real-time workloads and is accessible via the Snowflake Marketplace using Snowflake credits.

ThoughtSpot provides native support for Snowflake Semantic Views, allowing a seamless flow of Snowflake-defined business logic, metrics, and relationships into ThoughtSpot.

Industry context emphasizes grounding agentic AI in governed business context and data sovereignty to scale trusted AI, with Snowflake underscoring governance and consistency through Semantic Views and Spotter Semantics.

Spotter now integrates with Cortex Analyst and Cortex Agents, enabling Cortex-derived insights to be accessed directly within ThoughtSpot without switching systems.