Human brain is one of the most intricate organs which is responsible for multiple body functions and acts as an integral body organ for most bodily functions. It is made up of billions of nerve cells known as neurons. The neurons communicate with each other through trillions of connections called synapses. Neurological disorders like epilepsy, seizures, and other brain abnormalities require continuous, detailed and very sophisticated monitoring of human brain activity. Electroencephalography (EEG) is a very widely used non-invasive techniques for monitoring brain activity. The intricate nature of EEG data poses significant challenges, especially in distinguishing between normal and harmful brain activity in real time. In this study, a deep learning (DL) based framework to automate the tedious process of classification of EEG samples has been proposed. To achieve this, Convolutional Neural Networks (CNNs), which are a specialized type of deep learning architectures well known for their strong performance in image recognition tasks have been employed. In addition to the model, this system incorporates explainability techniques to ensure that the overall approach of decision-making is transparent and easy to interpret for medical professionals.