Abstract: A method comprises gathering, by a large language model (LLM), a plurality of static information from a plurality of wellbore reports; correlating, by the LLM, the plurality of static information with a plurality of dynamic information from a plurality of data sources; providing, by the LLM, a predictive analysis of a wellbore condition based on the gathering and the correlating; and retraining the LLM based on the predictive analysis.

Abstract: Some implementations include a computer-implemented method for determining one or more subsurface properties from one or more graphical oil field logs created during oil field of one or more oil field operation. The computer-implemented method may include graphically analyzing the graphical oil field logs; determining one or more regions of interest in the graphical oil field logs; generating, based on the graphical analysis of the graphical oil field logs, a respective textual description for each respective region of interest in the graphical oil field logs; and determining, by a large language model (LLM), one or more lithologies for each respective region of interest based on the textual description for the respective region of interest.

Abstract: Embodiments of a method, system and machine-readable medium for generating a summary of drilling and operation reports based on a user prompt are disclosed herein. In one embodiment, a method comprises receiving a user prompt that was parsed by a natural language processor; generating a first summary of a plurality of drilling reports for at least one wellbore based on a report repository created from the plurality of drilling reports from at least one wellbore; indicating a sentiment score of the first summary; and generating a final summary based on the first summary and the user prompt.

Abstract: Embodiments of a method, system and machine-readable medium for generating a summary of drilling and operation reports based on a user prompt are disclosed herein. In one embodiment, a method comprises receiving a user prompt that was parsed by a natural language processor; generating a first summary of a plurality of drilling reports for at least one wellbore based on a report repository created from the plurality of drilling reports from at least one wellbore; indicating a sentiment score of the first summary; and generating a final summary based on the first summary and the user prompt.