<p>In the last few years, considerable development has taken place in the practices of civil engineering through the application of artificial intelligence-based methodologies. The accuracy of the results obtained has improved, the computation costs are reduced, and the application of the methodologies has resulted in sustainability as well as optimization of the complex civil engineering systems. The present research work has proposed a novel hybrid computational framework wherein the application of Machine Learning (ML) as well as Multi-Objective Water Strider Optimization (MOWSO) has resulted in improved seismic behavior of Light Steel Keel Frame-Wall systems. The artificial neural networks are trained in the present research work to accurately simulate the nonlinear seismic behavior of Light Steel Keel Frame-Wall systems. The accuracy of the results obtained is reflected by the value of R^2 being greater than 0.98. The results obtained in the present research work have resulted in a reduction of inter-story drift by 35%, improvement of energy dissipation capacity by 28%, as well as a reduction of materials by 20%. The results obtained in the present research work are accurate enough to satisfy the code requirements, reduce computation time by more than 90%, as well as robust performance. The proposed approach establishes an efficient, adaptive, and sustainable paradigm for next-generation seismic design of lightweight steel systems.</p>