Energy giants cut data discovery time 30% with Cognite’s industrial AI platform

Industrial software provider Cognite showcased real-world customer deployments of its AI-powered data platform at its recent Impact event in Houston, demonstrating how major energy companies are achieving measurable results with industrial AI. Unlike many competitors making broad AI promises, Cognite presented concrete examples of customers using its Cognite Data Fusion (CDF) platform to reduce data discovery time by up to 30% and streamline operations across multiple facilities.

Key customer deployments: Four major energy companies are scaling Cognite’s platform across their operations with impressive results.

• NOVA Chemicals, Canada’s largest polyethylene producer, started with one plant and now aims to deploy CDF across eleven facilities by year-end, moving toward a single enterprise-wide knowledge graph.
• TotalEnergies, the French integrated energy company, is rolling out CDF across 39 assets over three years, combining data from over 200 production systems with initial focus on front-line worker health and safety.
• AkerBP, the Norwegian oil exploration firm, operates its newest asset, Yggdrasil, as a remote-first operation managed by just 2 onshore staff, using AI agents to cut root cause analysis from weeks (sometimes 9 months) down to hours.
• Cosmo Energy, the Japanese petrochemical firm, deployed CDF across all three refineries, managing 250 billion data points and 5 million engineering documents for over 1 million pieces of equipment.

The foundation technology: CDF serves as an industrial data operations platform that creates comprehensive knowledge graphs from disparate source systems.

• The platform combines data from different systems into consistent, structured industrial knowledge graphs—think of it as creating a master directory that connects all of a company’s scattered information.
• Cognite partners with Microsoft Azure, Databricks, and Snowflake to help customers manage data across multiple cloud systems without unnecessary duplication.
• AkerBP divides its cloud data strategically: operational data in CDF, IT and business intelligence in Microsoft Fabric, and subsurface data in Azure Data Manager for Energy.

Atlas AI makes complex data accessible: Cognite’s Atlas AI platform transforms sophisticated knowledge graphs into user-friendly interfaces for front-line workers.

• Products like InField give workers easy access to insights without needing to understand the underlying knowledge graph complexity—like having a smart assistant that knows everything about your facility.
• Atlas AI includes conversational chatbots for querying data and an AI agent workbench for creating new agents and automated workflows.
• AkerBP uses multiple agents working together to tackle complex problems like root cause analysis investigations.

Why this matters: Cognite’s approach emphasizes the foundational data work required for successful industrial AI implementation.

• The company targets firms willing to invest in proper data context, lineage, provenance, and semantics rather than seeking quick AI shortcuts.
• As analyst Paul Miller from Forrester notes, successful industrial AI requires understanding that “the shiny AI app that pleases the Board or impresses the prospective customer is a tiny little piece of the iceberg, sitting on top of a whole mountain of data ops hard work.”
• This methodical approach positions Cognite for customers seeking long-term, scalable industrial AI solutions rather than proof-of-concept demonstrations.