The study focuses on predicting the 28-day compressive strength of ordinary concrete based on 7-day strength using Artificial Neural Networks (ANNs). As a critical material in construction, accurately forecasting concrete’s compressive strength is essential for effective planning and quality control. Traditional methods, including linear regression, have limitations in capturing the nonlinear relationships between mix variables and resulting strength. By utilizing ANNs, which excel in identifying complex patterns, this research proposes a reliable model for compressive strength prediction. Experiments involved testing twelve concrete mix designs with a fixed 10% micro-silica replacement for cement. Each sample’s slump (workability) and compressive strength at 7 and 28 days were recorded, with data processed for ANN. The model achieved high accuracy, with a correlation coefficient (R) 0.99 between predicted and actual 28-day strength values. Micro-silica enhanced early and long-term strength, with mixes achieving up to 44.5 MPa at 28 days. Linear regression analysis also yielded robust predictions (R²=0.958), underscoring the viability of statistical methods alongside ANNs. The ANN model enables early strength estimation, assisting construction decision-making and optimizing resource use. The findings support ANN and micro-silica integration for robust, high-performance concrete. This approach aligns with advancements in AI-driven construction practices, offering a practical tool for early-stage strength prediction and quality assurance in concrete applications.