Lucy in the Field: AI Joins the Shale Workforce

Baker Hughes and Expand Energy scale AI across thousands of shale wells, pushing production management into a new data-driven era

4 Feb 2026

A new phase in US shale production is under way after Baker Hughes signed a multi-year agreement with Expand Energy to deploy its Leucipa automated field production system across thousands of gas wells.

The rollout will focus on major gas basins including the Marcellus, Utica and Haynesville. The scale of the agreement signals a shift away from limited digital pilots towards wider adoption of data-driven production management across large asset portfolios.

Expand Energy describes itself as North America’s largest natural gas producer, giving the project significance at a time when operators face pressure to improve efficiency and maintain stable output. Producers have sought to standardise operations as drilling activity moderates and investors demand stronger returns.

Baker Hughes says Leucipa is designed to help operators prioritise field-level decisions across multiple domains. These include production optimisation, artificial lift, network performance and maintenance scheduling. The platform also incorporates oilfield chemicals management, which remains central to corrosion control, scale prevention and flow assurance in gas wells.

By bringing these areas into a single data environment, companies aim to align day-to-day operational choices with real-time production data rather than relying on periodic manual adjustments.

Under the agreement, Expand Energy will also pilot “Lucy”, described as the Leucipa AI Production Assistant. The generative AI interface is intended to provide conversational access to production data and surface operational insights for engineers and field teams.

Industry reports indicate that Leucipa will be delivered as a software-as-a-service platform hosted on Amazon Web Services, allowing integration across Expand’s well base and supporting scalability as deployment expands.

The broader direction across US shale is towards continuous, analytics-led management of large well populations. As operators embed AI tools into routine workflows, optimisation is increasingly framed as an ongoing process rather than a series of discrete interventions. That shift could influence how companies manage chemicals usage, maintenance planning and operational risk across mature gas fields.