Inference of Tacit Meeting Knowledge via LLMs: Toward Emotion-Enriched Organizational Memory

Organizations risk losing critical tacit knowledge when meeting records capture only what was said rather than how it was expressed, such as emotional tone, tension, or unspoken dynamics. We investigate whether integrating tacit knowledge into Large Language Model (LLM)-based meeting summaries can enrich organizational memory. Following a Design Science Research approach, a prototype was developed to transform meeting audio into tacit-enriched knowledge outputs. Initial experiments explored audio-only Speech Emotion Recognition but revealed significant reliability issues in realistic, multilingual and interruption-rich environments, particularly Swiss German. Consequently, the solution pivoted toward a text-based LLM framework using structured, role-based prompt engineering to extract tacit insights from transcribed meetings. The artefact generated two summaries, one capturing explicit and the other capturing implicit emotional and tacit cues. The exploratory evaluation involved 32 participants across 17 workplace meetings and examined perceived accuracy, additional insight, usefulness, and ethical acceptability. Participants frequently perceived additional contextual insight beyond conventional factual summaries, while also expressing concerns about unsupported interpretations and potential surveillance. As no independent ground truth or controlled comparison was used, the results provide evidence of perceived usefulness rather than verified improvements in organizational knowledge retention.