Abstract- Malware detection is a significant challenge in today's digital landscape. As new forms of malware are continuously being developed, traditional detection techniques often fall short due to their inability to detect these new strains. This paperintroduces meaningful features that effectively capture various types of malware, including viruses, worms, Trojans and Ransomware on Edge devices. The paper used a model that implemented Random forest classifier for feature selection and a support vector machine (SVM) model for Malware detection. Object-Oriented Analysis and Design (OOAD) methodology was used to as the design methodology, which involved identifying and modeling the different components of the system and their interactions. The system was developed using Python programming language, with an emphasis on model deployment via Python Flask for web-based testing and execution. The experimental results demonstrate the effectiveness of the proposed systems when compared with other existing system. The result gotten from proposed system is better than that of the existing system by achieving a detection accuracy of 99.98% which is better than existing techniques. This dissertation presents a promising direction for improving malware detection using support vector machine (SVM) model and highlights the potential for collaborative learning approaches to overcome the challenges of traditional centralized approaches. This result simulates edge device that performs malware detection. It measures the latency for each detection and prints whether the latency is high or low. After the simulation, it plots a graph to visualize the latency over multiple requests. Which shows that the proposed model had low latency between 0.25secs to 0.15 secs on multiple requests.