Thomson Reuters is building its enterprise AI and data platform on Snowflake, using it as a single data foundation across more than 37,500 governed tables and 350 data sources.
The move is designed to give Thomson Reuters one governed source of data for internal teams working across legal, tax and regulatory information. More than 1,500 internal users, including data engineers, analysts and business leaders, use Snowflake daily to access data and generate insights.
Thomson Reuters chose Snowflake in 2021 as it sought to bring data governance, security and infrastructure together on one platform. That work now underpins the company's internal My Data Space platform, where central teams build and share data products across the wider organisation.
The setup supports AI development as well as data management. Thomson Reuters is using Snowflake Cortex AI to work with regulatory data and produce near real-time insights, while Snowflake CoCo helps teams move legacy systems onto Snowflake.
According to the companies, the shift is already affecting performance in products including CoCounsel and Westlaw as Thomson Reuters consolidates supporting data pipelines. Some workloads are running up to 3.4 times faster, and analysis that previously took weeks now takes seconds.
The changes have also reduced manual data preparation in key workflows, allowing teams to move away from static reporting and towards more immediate analysis for business decisions.
For Thomson Reuters, the emphasis is on using AI in areas where accuracy and control matter because its end users work in regulated, high-stakes professional settings. The company described its governed data structure as a necessary foundation for developing AI tools that professionals can trust.
"With Snowflake Cortex, we're accelerating how we build and scale AI across Thomson Reuters," said Caitlin Halferty, Head of Data & Analytics, Thomson Reuters.
"For us, the real value is not just speed. It is the ability to innovate in a governed environment where our teams can turn complex regulatory data into actionable insights while maintaining the trust, control and reliability required for high-stakes professional use," said Halferty.
Snowflake said the Thomson Reuters deployment shows how companies are trying to combine AI rollout with tighter control over data handling, particularly in sectors where outputs must stand up to scrutiny. That has become a central issue for organisations adopting generative AI tools while managing legal, compliance and audit requirements.
Operational shift
The scale of the Thomson Reuters data estate illustrates the operational challenge. Bringing together 350 data sources and tens of thousands of governed tables requires standardised access and oversight across multiple business functions, while still giving users enough flexibility to run analysis and build applications.
In practice, the company's approach reflects a wider push among large information providers to reduce fragmented internal data estates. Rather than keeping separate systems for analytics, applications and AI work, companies are increasingly trying to place those activities on a common platform that can be monitored and controlled centrally.
That trend is particularly relevant for Thomson Reuters because much of its business serves professionals who rely on current and defensible information. Legal, tax and regulatory products depend not only on access to large volumes of data, but also on confidence in how that data is managed and used in AI-assisted systems.
Snowflake has positioned Cortex AI as a way for customers to use AI functions on governed data already held in its environment. For Thomson Reuters, that means using existing stores of regulatory and professional data without moving them into disconnected development workflows.
The use of Snowflake CoCo points to another part of the strategy: updating older technology systems while trying to avoid disruption to security and compliance controls. Many large companies face a difficult migration path when moving legacy workloads into newer data environments, especially where existing systems support established commercial products.
Christian Kleinerman, Executive Vice President of Product at Snowflake, said the Thomson Reuters project showed how governance and AI deployment can be handled together on one platform.
"Thomson Reuters is demonstrating how enterprises can scale AI and governance together on a single platform," said Christian Kleinerman, Executive Vice President of Product, Snowflake.
"By building on Snowflake, they're creating a trusted foundation that allows teams to move faster and scale AI across the business," said Kleinerman.