Intelligence chat copilot app for well-log processing and analysis

Marcelo Guarido, Kristopher A. Innanen

AI assistants powered by large-language models have been applied to different fields for quick analyses, data summarization, and advice chatbots. We proposed an ensemble of AI assistants for an app for Well-log analysis and data modelling. It works as a chat assistant that can access SQL databases to extract data and process it to return a table, a chart, or a machine-learning model based on the user's request. Our tests showed that the app could yield satisfactory results for the three output requests, but it required trial and error in asking the question or request. This application can create complex modelling and analyses for wireline logs without programming, making it accessible to a more significant number of geoscientists.