Abstract The oil and gas sectors face significant challenges in managing vast, disparate geoscience datasets. This project addresses these challenges by implementing a centralized solution for subsurface data and documents, providing users with rapid access to data through web-based tools. Generative Artificial Intelligence (AI) offers conversational natural language queries across unstructured data and augments visualization with summaries of complex subsurface data environments. The solution integrates a unified subsurface data lake with standardized metadata tagging and a custom-trained natural language processing (NLP) engine specifically fine-tuned for geoscience terminology. Secure API connections provide comprehensive access to well databases, seismic repositories, and technical reports. Early adoption has demonstrated transformative benefits, including a 70% reduction in time spent searching for information and a 60% decrease in technical data management help desk requests related to data location. The intelligent data access interface lowers the barrier to accessing and understanding subsurface knowledge, empowering teams to uncover insights from previously disconnected information. By converting complex data retrieval into a simple, conversational process, this project demonstrates how AI-powered search can fundamentally change how geoscientists interact with technical data, turning raw information into actionable knowledge at the speed of conversation. This approach aligns with industry trends highlighted by firms such as Gartner and McKinsey, which emphasize the importance of breaking down data silos and using generative AI to automate knowledge work and enhance decision-making.
Nor et al. (Mon,) studied this question.