This article shown the performance evaluation of AISI 4340 using the hot-turning method. AISI 4340 is a medium carbon steel known for its impact resistance and is commonly used in gear components of airplane systems, crankshafts and connecting rods. This research focuses on developing mathematical models using Artificial Neural Networks (ANN) to analyze the relationship between dependent and independent variables. The ANN is suitable for nonlinear machining operations with high accuracy and without a predetermined model. Modelling techniques, specifically ANN, are suggested to improved machining effectiveness and reduce production time and costs. This article explains the architecture and function of ANN, describing neurons, weights, and activation functions. This research methodology includes data pre-processing, network training using Back Propagation (BP) algorithm, and selecting activation functions. The Levenberg–Marquardt algorithm is used for training, and the hyperbolic tangent sigmoid transfer function is utilized as the activation function. The results show that a 4–17-1 network structure provides the closest prediction to experimental data for surface roughness, with an MSE of 0.004393106, and a prediction error of 3.79522%. The result demonstrates usefulness of ANN in predicting the machining quality.