This paper proposes an integrative approach to Grounded Theory Methodology (GTM) that combines human expertise with large language models (LLMs) to advance qualitative research in Information Systems (IS). Addressing the limitations of current computational GTM methods, which lack the ability to generate contextually rich explanations alongside codes, we leverage LLMs’ generative capabilities to enhance theory construction from text data. Grounded in the pragmatist perspective, which emphasises action and its consequences, we compare results from our approaches and aim to offer insights on how LLMs (integrated with humans) perform across the GTM stages.