Autism Spectrum Disorder (ASD) is a complex neurodevelopmental condition characterized by challenges in social interaction, communication, and repetitive behavioral patterns. Early and accurate diagnosis of ASD is crucial for timely intervention and improved quality of life. Traditional diagnostic methods rely heavily on clinical observation and behavioral assessments, which are often time-consuming and subjective. This paper proposes a machine learning-based system for the early prediction of Autism Spectrum Disorder using behavioral and demographic features. The system evaluates key screening indicators derived from the AQ-10 questionnaire and applies multiple classification algorithms including Random Forest, Decision Tree, Support Vector Machine (SVM), and Logistic Regression to identify ASD patterns. The best-performing model is selected and integrated into a web application built using Flask, enabling real-time prediction through a user-friendly interface. The proposed approach achieves a classification accuracy of 93.6% and aims to assist healthcare professionals in the early screening and diagnosis of ASD.