An approach to managing a construction organization based on the use of artificial intelligence methods

The article examines an approach to managing a construction organization based on the application of artificial intelligence methods, driven by the need to overcome systemic industry problems: staff shortages, low labor productivity, high uncertainty and risks in the implementation of investment and construction projects. The analysis of existing theoretical approaches to construction organization management is carried out, identifying interactive, organizational-technological, system-digital, process approaches and cybermanagement, which revealed their limitations in the context of digital transformation of the economy. An approach to managing a construction organization is proposed, containing four levels (strategic, tactical, operational and executive) and artificial intelligence methods: large language models (LLM) for market analysis and generation of strategic alternatives; neural network forecasting for optimization of schedules and cost of construction and installation works; computer vision and IoT for construction site monitoring, occupational safety and quality control; AR systems and digital twins for visualization and support of executive operations. The proposed approach, based on the "human-in-the-loop" principle and a unified information space of digital twins, ensures self-learning of models and continuous improvement of the accuracy of management decisions in conditions of high uncertainty in construction production.