Distributed Denial of Service (DDoS) attacks pose a critical cybersecurity threat, capable of disrupting online services and causing operational challenges. In this work, we apply machine learning techniques, specifically Deep Neural Networks (DNNs), to analyze and detect DDoS attacks using a Software-Defined Networking (SDN)-specific dataset. The dataset contains 104,345 instances with 23 features, where the output feature classifies the traffic as benign or malicious. After preprocessing steps such as handling missing values, one-hot encoding, and normalization, the dataset was reduced to 103,839 instances with 57 features. The proposed DNN model achieved an accuracy of 99.38%, which is approximately 1.21% higher than the next best baseline model, XGBoost, which recorded an accuracy of 98.17%. Furthermore, real-time data processing was implemented using MongoDB and a scalable backend, enabling continuous updates and attack detection, with results visualized on a dashboard for efficient monitoring. These findings highlight the potential of deep learning models for detecting and mitigating DDoS attacks in real-time.