In the present-day scenario of complex urban infrastructure, studying pile raft–tunnel interaction is essential to ensure safe, efficient foundation performance amid increasing subsurface conflicts and variable soil conditions. Reliability-based design is crucial today as it quantitatively assesses risk and safety under uncertainty, a factor not explicitly covered in the current Indian code of designs. This chapter presents a reliability-based design approach for a combined piled raft foundation system influenced by an underlying tunnel. The analysis accounts for key sources of uncertainty by considering the variation in undrained shear strength (Su), soil stiffness (Es), tunnel diameter (D), and tunnel depth (Y) modelled in PLAXIS 2D. For this purpose, the project utilized the serviceability limit state concept with settlement as a primary measure of performance. To minimize the high cost of running an infinite number of finite element analyses, surrogate models were developed based on the responses obtained from the PLAXIS analysis. Various surrogate models such as Multivariate Adaptive Regression Splines (MARS), Artificial Neural Network (ANN), Random Forest (RF), Support Vector Regression (SVR), etc., were developed and tested for their performance metrics. The best surrogate model for this purpose is the ANN model, as it resulted in the lowest prediction errors and the highest value of R-squared. The ANN model was later employed for the reliability analysis using First-Order Reliability Method (FORM), Monte Carlo Simulation (MCS), and Mean Value First-Order Second Moment (MVFOSM) methods for evaluating the chances of settlement exceeding the permissible limit.