In today’s knowledge-oriented society, the availability of high-quality information is of paramount importance for businesses. Due to the uncontrolled growth of data in businesses over the last decades, it is nearly impossible to provide a high-quality knowledge base for users that want to make informed decisions. In this setting, the undocumented context creates further problems for intelligent search algorithms by providing out-of-context results due to the presence of ambiguous entities in the knowledge base. This justifies the need for an IT system that provides intelligent functionality to enable quality-assured growth of the enterprise knowledge base and better support user needs. An essential feature to assure high quality is to involve the user in the evolution process of the knowledge database. This requires new concepts that users can apply in practice. In this thesis, we design a system that supports the quality-assured growth of the company’s knowledge base by linking new entities to existing ones and providing an intelligent natural language user interface that takes full advantage of the linked enterprise data. It is also capable of disambiguating entities on-the-fly (Disambiguation on Write) as user input is being entered, and provides query capabilities on top of the knowledge base to provide high-quality results. We discuss how the concept can be implemented using various technologies from the field of Natural Language Processing and Semantic Web and show a small web-based prototype. We define two use case scenarios from the domain of manufacturing and data science that demonstrate the power of our concept and then evaluate our prototype against the different functional and non-functional requirements that arise from these use case scenarios. Finally, we discuss various technical challenges that we faced during the implementation of the prototype and identify ideas for future work that would further improve the concept.