Oracle on Tuesday said it has partnered with Snowflake to make the latter’s AI Data Cloud available as a native service on Oracle Cloud Infrastructure (OCI).
The partnership, announced at Oracle’s ongoing AI World conference, aims to help enterprises reduce data movement between the two platforms and the associated costs while accelerating analytics and AI workloads, the companies said in a joint statement.
That means enterprises with data stored in Oracle Fusion Cloud Applications, NetSuite, Oracle Autonomous AI Database, MySQL HeatWave, and Oracle NoSQL Database will be able to access Snowflake’s data analytics and AI capabilities on OCI without having to move their workloads to another environment, they added.
Integration could increase productivity and scale up AI adoption
The partnership could help reduce costs associated with data movement for enterprises, analysts say.
“Large Oracle application customers that already use Snowflake stand to benefit first, particularly those moving ERP and finance data across clouds every night. Keeping high-volume finance and supply chain data within OCI can eliminate cross-cloud data transfer charges, making egress savings the clearest and most immediate benefit,” said Ashish Chaturvedi, executive research leader at HFS Research.
“NetSuite customers are also likely to experience similar benefits as they hold valuable Oracle data and have little engineering capacity, so a native path into Snowflake removes integration work they could rarely staff,” Chaturvedi added.
For enterprise development teams, the same reduced data movement could mean spending less time building pipelines to move supported Oracle data into Snowflake and more time building applications that use it, said Stephanie Walter, practice leader of AI stack at HyperFrame Research.
Data professionals, meanwhile, could spend less time maintaining the extract-and-load jobs needed to move data between the two platforms, freeing them to focus on higher-value analytics and AI projects, Chaturvedi said.
Similarly, for CIOs, the same reduction in data movement and integration complexity could help address data readiness and cost challenges that often slow enterprise AI initiatives, Chaturvedi added: “Making ERP data available to AI tools faster and simplifying procurement through a single contract and bill could reduce some of the barriers to adoption.”
Data governance challenges remain
Running both services on OCI, however, does not automatically eliminate the work involved in integrating enterprise data.
Developers will still need to manage integrations, permissions, and data quality to ensure applications can access and use the right information, particularly when working with data sources outside Oracle, Walter said.
For data professionals, the integration also does not eliminate data duplication, since data brought into Snowflake still exists as a copy.
Instead, it introduces operational complexity through the need to manage two governance catalogs, separate cost models, and potentially multiple Snowflake accounts across cloud environments, said Michael Leone, principal analyst at Moor Insights and Strategy.
That means enterprises will also need to establish which version of a business metric should be treated as authoritative, particularly when data is governed across multiple systems, echoed Chaturvedi.
For CIOs, these challenges raise broader questions about cloud portability, commercial flexibility, and governance, Chaturvedi pointed out.
Therefore, they will need to assess whether Snowflake data and code can be moved to another cloud without incurring high costs, how easily committed credits can be allocated between OCI and Snowflake as business priorities change, and whether governance policies remain consistent across Snowflake deployments on different clouds, Chaturvedi advised.
For enterprises whose data environments are already centered on Oracle, the integration could make sense, but organizations whose data estates are primarily on AWS or Microsoft Azure could find that adding OCI introduces another environment to manage, increasing operational overhead and potentially adding to technical debt, he further said. Both companies are yet to confirm when the offering will become generally available to customers.
