Snowflake and Lumen plan private network path for on-premises AI data

The network as the bottleneck

Snowflake argues that for many organizations, critical business data never reaches the cloud. It stays in data centers, branch offices and edge locations, where it supports local operations but is invisible to analytics and AI projects.

The problem is most acute in regulated industries such as healthcare, telecom, financial services and government. Snowflake says their compliance rules govern where data resides, who can access it and how it is protected. Those rules do not necessarily prohibit cloud use, but they can limit which cloud environments are allowed and require specific safeguards and approvals. For certain jurisdictions or categories of sensitive or classified data, restrictions may rule out any transfer over the public internet.

According to the company, public internet connectivity often fails to meet the security needs of these sectors, while private connectivity can demand complex configuration and lengthy setup. Moving on-premises data into a cloud platform typically requires months of provisioning and coordination across data, security and network teams. The result, Snowflake says, is a gap between the data an organization has and the data its AI can actually reach.

How the Lumen and Snowflake path is meant to work

Under the partnership, Snowflake's DCP will be available to run as a container image on Lumen infrastructure. It is meant to establish a secure connection between a customer's distributed data environment and their Snowflake account. The planned data path uses Lumen's private network, and according to Snowflake, no data bytes are ever exposed to the internet.

The workflow combines DCP's connectivity path with Snowflake Openflow, which handles data integration and movement from supported sources into Snowflake. Once the data is in Snowflake, Cortex AI agents can query it alongside cloud sources.

Snowflake gives several example use cases:

• Health systems can replicate Electronic Health Record, imaging and lab data into Snowflake using Openflow to build readmission-risk models while enforcing row-level security.
• Banks can feed transaction and branch data into fraud detection.
• Telecom operators can combine edge network data with billing information to predict outages and churn.

Snowflake says private connectivity provisioning has historically required five to eight months of lead time. The Lumen-enabled path is designed to deliver preconfigured access upon customer setup, without the lengthy coordination between networking, security and data platform teams that enterprises have come to expect.

What the companies and an analyst say

Artin Avanes, Head of Core Data Platform at Snowflake, said: "Agents are only as good as the data they can reach, and today, too much of that data is stranded in locations the cloud can't easily access." He added that the companies share customers across financial services, telecom, healthcare and the public sector.

Jim Fowler, Chief Technology and Product Officer at Lumen Technologies, said the two companies "intend to simplify the connectivity needed to make distributed data more accessible for analytics and AI, while supporting customers' governance and security needs."

Sanjeev Mohan, Principal at SanjMo, described the network as one of the most overlooked challenges in enterprise AI. He said that if the intent delivers, "More of an enterprise's real data becomes reachable by its AI, without trading away governance."

What it means for customers

The announcement is framed in forward-looking terms. Snowflake describes the capabilities as planned and notes that the article does not commit the company to delivering any product offerings. The source gives no pricing, availability dates or regions.

For regulated enterprises, the pitch is that private connectivity becomes part of the data architecture rather than a months-long provisioning project. Once distributed data sits in Snowflake's AI Data Cloud, teams can use it alongside existing business data, with centralized security and governance over who can access sensitive data and how it is used. Snowflake directs interested customers to contact their account team to learn more about the offering.