In this session, Chad Harkness (Chief Revenue Officer at Silicon Valley startup Geminus AI) and Amanda Fagnou (Director of Strategy and Marketing for SLB's digital division, formerly Schlumberger) explore how artificial intelligence is being used to control mission-critical industrial assets in the oil and gas sector — an industry with roughly $1.3 trillion in annual spend, where every decision can carry irreversible consequences.
Amanda Fagnou traces the evolution of digital technology in the industry — from mainframes to the cloud, and now to edge computing for operations — and outlines the two conditions SLB considers essential for AI to work in this sector: reliable, accessible data, and specialized "foundation models" trained on sector-specific data rather than generic LLMs alone.
Chad Harkness then presents Geminus AI's approach, built around "neighborhood models" — models that combine physics-based, historical, and real-time data to understand an industrial system in all its specificity, where large generic models or emerging "world models" remain too general. The session closes with concrete deployment examples, including refineries and an oil production network in Bahrain.
