Cognite: Databricks & Snowflake partnerships forge new links in industrial AI value chain

What developers want

Cognite reminds us that data science and AI-focused data engineering teams require well-governed, high-quality data to train and execute models that solve complex field-level use cases.

“Our joint customers are seeking a governed, open AI and data foundation. By integrating the Databricks Data Intelligence Platform, including products like Agent Bricks, and the Cognite Industrial AI and Data Platform, we’re empowering them to innovate without compromise,” said Andy Kofoid, president of global field operations at Databricks. 

Why industrial AI is tough

Industrial AI solutions require real-time, accurate data with context for reliable results; however, this is often hindered by raw, complex and siloed industrial data. Cognite and Databricks say that they aim to directly address this by utilising semantic models (knowledge graphs) to provide the necessary enriched information, referred to as context, on the customer’s data within their environment, thereby moving beyond generic data models.

This work aims to enable access to high-quality, trusted data required to build domain-specific AI agents and applications that reliably solve industrial challenges. It is hoped to eliminate expensive data duplication, storage and complex ETL pipelines, allowing engineering teams to focus on high-value AI.

“Customers deserve the freedom to choose the best platforms to meet the specific needs of their mission-critical operations. This partnership with Databricks is a commitment to an open ecosystem, where we use open standards and close collaboration to tear down data silos and deliver best-in-class results,” said Girish Rishi, chief executive officer at Cognite. “By bringing together two industry-leading platforms, we’re not just sharing data; we’re fundamentally multiplying the value our joint customers can derive, accelerating their journey to achieve significant, measurable results from their AI investments.”

The companies say that are working to harness open standards (such as Delta Sharing and Apache Iceberg) for cross-platform connections, eliminating vendor lock-in and allowing users to combine the best tools for data, analytics, and AI.

Snowflake unity

What’s the point in one data partnership for enterprise-wide industrial AI at scale when you can have two? Cognite thinks that’s a silly proposition, of course it should align its work with Snowflake as well.

These days styling itself as the “AI Data Cloud” company, Snowflake will now form a strategic partnership (once again with plans for a bidirectional, zero-copy data sharing integration) that unites the Cognite Industrial AI and Data Platform and the Snowflake AI Data Cloud. 

This collaboration is promised to offer what the firms grandly call a “unified single source of truth for industrial intelligence” across an enterprise (which in this case Cognite specifies as everyone from field operators to executives) to drive operational efficiency. This is because commercially-aligned business analysts require the same data as industrial operators to solve complex use cases in the field and build solutions that drive cost efficiencies. 

Industrial AI solutions with the same challenges mentioned above are being addressed by Cognite and Snowflake with what the companies say is a unified foundation for easy access to intelligent industrial data for users across the enterprise.

Bidirectional flow of AI-ready data

The integration will utilise a zero-copy data sharing integration to deliver a bidirectional flow of AI-ready data between the Cognite Industrial AI and Data Platform and the Snowflake AI Data Cloud. 

“Cognite is a prime example of the power of our connected ecosystem, providing the essential domain specialisation needed to transform complex industrial data into high-quality, AI-ready intelligence,” said Kieran Kennedy, global vice president, Snowflake Data Cloud product partners. “By directly leveraging the data where it resides, this partnership securely and seamlessly makes Cognite’s real-time, AI-ready industrial data available to consumers in Snowflake.”

This enables Snowflake end-users across the enterprise to gain real-time access to unified, domain-specific industrial data needed to power AI solutions and agentic workflows, while simultaneously allowing insights derived by these users to continuously enrich the Cognite platform. This ensures that all stakeholders have trusted, unified, and timely industrial intelligence, resulting in greater operational impact.

“Snowflake is one of the most important and widely adopted platforms in the enterprise, and this partnership allows enterprise analysts to leverage the intelligent industrial data to solve complex, cross-functional business problems,” said Trygve Rønningen, head of global revenue and partner enablement at Cognite. “We are working closely with Snowflake to ensure the effective sharing of industrial data through open standards to drive cost efficiencies across the entire organisation. This is how we unlock AI at scale and deliver more value to our joint customers.”

The open ecosystem here will enable users to use Snowflake’s Secure Data Sharing and open standards for data exchange.

Girish Rishi, chief executive officer at Cognite.