[DigitalToday reporter Chi-gyu Hwang] Cloud data platform Snowflake is accelerating a strategy to process AI work inside its own platform. Its $200 million partnership with OpenAI, its acquisition of Observe AI and the unveiling of Project Snowwork can also be seen in that context.

The Register reported on March 23 that Gary McConnell (게리 맥코넬), chief executive of Snowflake partner Virtuity, explained: "The OpenAI partnership, the Observe acquisition and Project Snowwork are pointed in the same direction. It is not about sending data out to AI, but about having AI work itself take place inside the data platform." He added: "In the past, companies had to stitch together a data warehouse, a feature store, a dedicated repository for preprocessing and storing and managing data, and a separate AI and ML environment, but Snowflake is integrating this structure into a single stack."

Snowflake's number of customers rose 70 percent over three years, from 7,800 in January 2023 to 13,330 in January this year. Large corporate customers under the Forbes Global 2000 also increased over the same period, from 573 to 790. Those large companies accounted for 43 percent of revenue of $4.7 billion in the most recent fiscal year.

Snowflake partnered with Google early this year and integrated the Google Gemini model into its AI inference service CortexAI. In February, it signed a $200 million agreement with OpenAI.

It also unveiled Snowflake Postgres, a database aimed at AI agents. Snowflake Postgres is based on the open-source Postgres database and supports using PostgresDB in its data lakehouse environment.

Snowflake has also recently started beta testing Project Snowwork. Project Snowwork uses role-based AI personas to provide preset workflows and metrics for departments including finance, sales, marketing and operations.