Malaria is one of the deadliest diseases on the planet and a significant challenge for the health departments. Despite efforts to eradicate this disease as it slumbers always in people’s lives, it still continues to kill millions through transmission by female Anopheles mosquitoes with the responsibility of Plasmodium parasites. Traditional methods like manual microscopy or using rapid diagnostic tests have been proven to have much human error that drives the need for more accurate detection and more efficient means of diagnosis. Custom convolutional neural networks (Hatami et al. in Classification of time-series images using deep convolutional neural network. SPIE, 2017 [1]) were presented here, and they were to work on identifiable healthy blood and infected blood samples to provide cost-friendly methods for detection at speeds faster than old methods. The model that was proposed, constituting three convolutional layers and fully connected layers, had amazing accuracy by using very limited computational resources. The CNN classifier achieved an accuracy level of 94.33% after going through training and testing over a range of blood sample images. It can be effectively inferred that by implementing deep learning (Vadavalli and Subhashini in Int J Recent Technol Eng, 2019 [2]) techniques in this context, not only does malaria (Dong et al. in Evaluations of deep convolutional neural network for automatic identification of malaria infected cells [3]) detection enhance but the approach also has a platform on the diagnosis of other diseases, thus marking a significant improvement of health globally.