<p>This work proposes a framework for monitoring and predicting air quality using the Internet of Things (IoT) and Machine Learning features. It features a four-layer architecture based on IoT four-layer architecture encompassing sensing, network, processing, and application layers. The methodology involves collecting data, preprocessing it, creating a prediction model utilizing Long Short-Term Memory (LSTM) for air quality, acquiring real-time data through sensors, and incorporating an intuitive web application for real-time air quality monitoring. Information gathered from multiple sites across India concerning pollutants such as PM_2.5, PM₁₀, SO₂, CO, NO₂, NH₃, O₃ and NO was used as input to the LSTM model, enabling accurate predictions of air quality indices. The framework incorporates various sensors connected to the ESP32 microcontroller and the Raspberry Pi microcontroller acts as a gateway, employing the Message Queue Telemetry Protocol for data transmission towards the cloud. The assessment of model performance is performed using metrics such as Mean Absolute Error and Root Mean Square Error, highlighting the significance of achieving high prediction accuracy and robustness. A web interface was developed to improve user interaction, providing real-time air quality data, visual trend analyses, and risk predictions. The proposed system attained a prediction accuracy of 80.75%, underscoring its capability to enhance awareness of environmental health risks via continuous monitoring and data-informed insights.</p>