Nowadays, air and noise pollution pose the greatest threat to the environment and to public health. Air and noise pollution are rising rapidly. In order to get it under control, close observation of it is strongly advised. This research addresses this problem by unveiling a method of detecting both ambient noise and the presence of potentially dangerous chemicals. Pollution is increasing at an alarming rate, and it’s starting to cause problems for living things. Pollutants like loud noises and hazardous gases have a negative impact on people’s health and want special attention. In this research, the proposed approach suggests the development of a network of sensors to monitor the environmental noise and air quality. A NodeMCU equipped with an ESP8266 WLAN adaptor is part of an Internet of Things (IoT) based Air and noise pollution monitoring system, together with a DHT11, a MQ9 gas sensor, and a Lm393 sound sensor. This monitoring system was developed with the help of neural network–based algorithms like CNN, SVM, and ELM to accurately forecast the level of pollution in a given area. Root-Mean-Square-Error (RMSE), Mean Square Error (MSE), and Mean-Absolute-Error (MAE) were used as performance indicators. Based on the performance, the Extreme Learning Machine (ELM) based training algorithm for single hidden layer feed forward neural networks (SLFNs) has revolutionized SLFN development. The results showed ELM as the most effective solution. The proposed model can now be used in high-pollution and noisy locations.