Air pollution is a global crisis with profound implications for public health and environmental sustainability. In addressing this issue in Chennai, Tamil Nadu, a novel Hadoop-based real-time air pollution prediction system is proposed. This research offers accurate air quality information for specific Chennai regions, aiding decisions and mitigating pollution risks through big data analytics and deep learning for air quality prediction. To expedite air quality prediction, a vast air pollution dataset is rigorously analyzed using a Hadoop-based analytics model. Specific locations in Chennai, including Perungudi, Royapuram, Manali, Alandur, Arumbakkam, Kodungaiyur, and Velachery, are assessed for upto- date air quality evaluations. The core of the research revolves around implementing deep learning models—Long Short-Term Memory, Convolutional Neural Network, and a hybrid Long Short-Term Memory-Convolutional Neural Network model. These models are trained to forecast AQI for selected areas over the next five years, with the hybrid model emerging as the standout performer, achieving 99% of accuracy rate and mean absolute error, mean square error, root mean square error rates of 7.95, 101.71, 9.65. This high accuracy and low error rates empowers policymakers and environmental agencies to make informed decisions, fostering healthier living conditions in Chennai. The integration of big data analytics and deep learning, promises improved air quality management in urban areas globally, addressing similar environmental challenges and enhancing overall quality of life.