Sayari, a commercial risk intelligence company, announced it has selected Snowflake (NYSE: SNOW) to rebuild its Commercial World Model, a database of company ownership, trade relationships, and corporate records drawn from more than 250 jurisdictions.

The migration consolidates 12 billion records from 715 sources into a single platform. According to the company, previous extraction methods could not capture up to 75% of the useful information contained in its archive of roughly 1 billion original documents, including corporate filings, government gazettes, and shipping records. Sayari said advances in AI, combined with the Snowflake platform, allow it to process that previously inaccessible content across the entire archive.

Sayari's engineering team used Snowflake CoCo, Snowflake's AI coding assistant, to assist with the migration. The company projects the move will reduce its data infrastructure costs by more than 50%.

"Rebuilding the Commercial World Model on Snowflake lets us read all of it at global scale," said Jessie Abell, chief product officer at Sayari. "Later this year, we will introduce products built on this foundation that deliver finished answers customers can trace, verify, and defend."

The Commercial World Model covers records in more than 20 languages and scripts, collected directly from official sources including corporate registries, customs bureaus, and regulatory agencies. Sayari said much of the data consists of deep web records not indexed by search engines and, in some cases, no longer available from their original publishers.

Sayari said it plans to introduce new products built on the rebuilt platform later in 2026. The company's existing customers include U.S. Customs and Border Protection and U.K. HM Revenue & Customs, according to the press release.