The loss of tacit knowledge, constituting the majority of organizational knowledge, significantly impairs an organization’s ability to compete. Traditionally, research and practice have focused on preventing knowledge loss through human- and technology-centered strategies. Nowadays Generative Artificial Intelligence (GAI) is disrupting many industries and brings a great paradigm shift for knowledge management. In particular the Retrieval Augmented Generation (RAG) capability emerges as a promising solution to combine the world knowledge of Large Language Models (LLMs) with domain-specific knowledge of companies.

This thesis explores how GAI with RAG capabilities supports knowledge retention and use by following a Design Science Research (DSR) methodology. Using multi-qualitative semi-structured interviews the problem is identified in a practical setting of the purchasing department of the company Bosch Power Tool Accessories. Based on an extensive literature review two solution setups for the explicit and tacit knowledge are suggested and developed. Findings from the evaluation confirm GAIs ability to support retention and use by retrieving and enriching codified organization relevant knowledge. While GAI with RAG capabilities shows promise in codifying expert insights and reducing organizational knowledge loss, its long-term effectiveness, particularly in capturing all facets of tacit knowledge, remains partially unresolved. This thesis contributes to the underexplored intersection of GAI and knowledge retention and provides a foundation for future empirical research and practical application.