Aker BP scales Cognite Atlas AI agents to automate workflows and cut analysis time 70%

Aker BP scales Atlas AI to streamline operations and cut costs

September 17, 2025 - Aker BP is expanding its use of Cognite Atlas AI across industrial and business workflows. The goal: higher efficiency, lower operational costs, and faster execution-without adding headcount friction.

Atlas AI pairs pre-configured agents with a low-code workbench, integrated directly with Cognite Data Fusion. For operations teams, that means faster build cycles for AI agents, fewer data silos, and more consistent decision-making.

Where AI is moving the needle

The company is automating document processing to speed up supply chain tasks and strengthen proactive barrier management. Engineers get quicker access to context and expertise, which shortens critical paths across routine and non-routine work.

"Our collaboration with Cognite on Atlas AI represents a significant leap forward in our digital transformation," said Aker BP chief digital officer Paula Doyle. "This partnership supports a continued AI-first, data-driven approach with agentic AI to drive efficiency, enhance decision-making, and optimize operations."

Early results are clear. Automated equipment registration aims to materially reduce engineering hours. New AI agents for root cause analysis identify relevant information, build cause maps, and recommend next steps-cutting engineer time on this analysis by more than 70%.

The initiative is part of a broader push to build an AI-ready data foundation for predictive maintenance and improved production chemistry. "By merging Aker BP's deep industry expertise with Cognite's Atlas AI, we are equipping them with the internal competence to scale AI across their entire business, fundamentally empowering their workforce for tomorrow," said Cognite strategic projects EVP Francois Laborie.

Separately, Aker BP recently completed the Omega Alfa exploration campaign in the Norwegian North Sea, resulting in an oil discovery.

What operations leaders can implement now

• Prioritize high-friction workflows: start with document-heavy processes, barrier management, and equipment registration.
• Tie agents to a clean data backbone (e.g., a unified operations data layer) to avoid duplicate logic and rework.
• Define hard metrics up front: cycle time, engineering hours saved, MTTD/MTTR, backlog burn-down, and permit-to-work lead time.
• Pilot root cause agents on recurring events; standardize cause maps and corrective action templates for reuse.
• Prepare for asset-level AI: condition data quality, maintenance logs, and chemistry datasets for predictive programs.

Learn more

Explore the companies involved: Aker BP and Cognite.

Building internal AI capability for operations? See AI courses by job role.