Deep learning is the advancement of machine learning which works on the training and prediction of complex data set. Deep learning offers valuable insights and improve decision-making processes when applied to meteorological data and reservoir operations. To identify patterns and trends, deep learning can process large volumes of environmental parameters like atmospheric pressure, wind speed, precipitation, temperature and humidity. Deep learning models are built using Artificial Neural Networks (ANN), which are designed to mimic the structure and function of the human brain. ANN networks process and analyze data on interconnected nodes (neurons). It automatically learn relevant features from the input meteorological data. These features can capture complex relationships between variables, such as the impact of temperature and rainfall on reservoir levels. Deep learning models are exposed to labeled historical meteorological and reservoir data during the training phase. They learn to recognize patterns and make predictions based on this information. Deep learning models use optimization algorithms to regulate the weights and biases of the neural network, gradually improving their predictive accuracy. Trained models can make predictions and forecasts based on real-time or future meteorological data. For example, they can estimate future reservoir levels based on weather conditions. Similarly, in urban area air pollution causes adverse effect on human health and environment hence monitoring the quality of air is now became a big concern. Air quality monitoring is the system to track, monitor, evaluate and measure the concentrations of different pollutants in the air and calculate Air Quality Index (AQI). After by examining AQI determine the air quality for that particular environment. The proposed AQM system trained on the standard available dataset on Kaggle which contains air quality data and air quality index (AQI) collected hourly and daily basis at various cities across the India. Real-Time IoT MQ135 sensor data collected by the hardware and through cloud send to ML model for AQI prediction. The real-time sensed values are displaed on the 16X2 display. If AQI exceed the set threshold value LED will blink and buzzer will give audible alert. The performance of the model evaluated for three different supervised Machine Learning models: Linear Regression, Decision Tree Regressor and Random Forest Regressor. The results shows that the model outperforms on Random Forest Regressor with RMSE value 1.21 and accuracy of 98%.