Tuffcrete concrete (ATC) has emerged as a promising material in modern civil engineering due to its enhanced durability and eco-friendly composition. This study presents the development of Artificial Neural Network (ANN) models to predict and optimize the compressive strength of Tuffcrete concrete based on experimental data. The dataset consists of 21 input features, including mix proportions (e.g., cement content, water-cement ratio, aggregate size distribution), material properties (e.g., tuffcrete polymer, slag content), and process parameters (e.g., mixing time, compaction level). The ANN models were trained and validated using these features to accurately forecast the compressive strength of Tuffcrete concrete under various conditions. The study demonstrates the model's ability to capture nonlinear relationships between input variables and compressive strength, achieving high accuracy metrics (e.g., R² and RMSE). Furthermore, optimization techniques were employed to identify the optimal mix design for maximizing compressive strength. Results reveal critical insights into the interplay between material properties and mechanical performance, paving the way for efficient mix designs tailored for specific applications. This work contributes to the advancement of machine learning applications in civil engineering, providing a robust framework for performance prediction and optimization of sustainable construction materials.