The digitization and optimization of active processes in the construction industry using methods of artificial intelligence is of strategic interest for those involved in construction. The aim of this project application is the development of an AI-based system for the semantic classification and indexing of all documents that are created and generated in a construction project. These documents (incl. plans) should be aggregated in clusters and visualized according to topic. By eliminating the manual categorization, keywording and clustering of information, the error rate is significantly reduced and efficiency increased. In addition, the AI system should only use natural search queries on the project platform to find and display relevant documents and create a short description from the „search result“. In this way, the user quickly and easily receives a compact and high-quality search result with all the relevant information. The automatic classification, keywording, content recognition and linking of the documents in a project eliminates a lengthy and time-consuming coordination phase for a project structure in advance. In order to achieve high quality and user acceptance of the automated processes, the AI models and methods used are specifically developed and optimized for the construction industry. To ensure a high and fast widespread effect, the AI-based solution developed here will be provided with open standard interfaces, so that integration into construction planning platforms and services established on the market WILL be easily possible. In this way, SMEs from the construction industry should also have easy access to modern, AI-based industry solutions.