Traditional building energy-saving (ES) methods usually only focus on individual improvements of building envelope structures and equipment systems, lack of comprehensive consideration of the overall system, and thus energy efficiency improvement is limited. This paper constructs a building ES algorithm based on the concept of system optimization, which aims to comprehensively improve building energy efficiency (BEE) and reduce energy consumption. It deeply analyzes the energy demand of the building and establishes an optimized design parameter set including key elements such as building layout, orientation, shading method and interior design parameters to ensure that the optimization process can cover important aspects of the entire life cycle of the building. Within the framework of Genetic Algorithm (GA), the BEC optimization problem is transformed into a multi-objective optimization task, and a fitness function is designed to balance multiple objectives such as BEE and cost, thereby effectively guiding the algorithm to find the optimal solution. The experimental results show that compared with the initial state of 180 kWh/ \(m^{2}\) /year, the energy consumption of the building after GA optimization is reduced to 120 kWh/ \(m^{2}\) /year, with an energy saving rate of 33.33%. The calculation efficiency value of the algorithm is 30 min. GA based on system optimization can not only significantly reduce BEC, but also maintain efficient calculation speed and accuracy. By comprehensively considering the interaction between various subsystems of the building, it can achieve an overall improvement in BEE.