Virtual Machine (VM) sizing refers to the process of selecting the appropriate configuration of computing resources—such as CPU, memory (RAM), storage, and network capacity—for a virtual machine to ensure that applications and workloads run efficiently, reliably, and cost-effectively. In cloud computing and IoT environments, VM sizing is particularly important because workloads can vary significantly depending on the number of connected devices, data processing requirements, latency constraints, and scalability needs. Proper VM sizing ensures optimal performance without over-provisioning resources, which would increase costs, or under-provisioning, which could lead to performance degradation and service failures. Since cloud platforms provide flexible resource allocation, VM sizing becomes a continuous process involving planning, monitoring, and adjustment based on workload demands. In the context of IoT systems, VM sizing plays a critical role because IoT applications typically involve large volumes of data generated from sensors, devices, and edge nodes. These workloads may include data ingestion, real-time analytics, machine learning inference, and storage management. Each of these tasks requires different resource profiles. For example, real-time analytics may require higher CPU and memory, whereas data archival may require more storage capacity. Therefore, VM sizing ensures that IoT services remain responsive, scalable, and cost-efficient while maintaining quality of service. There are two primary ways to perform VM sizing: vertical sizing (scale-up or scale-down) and horizontal sizing (scale-out or scale-in). Both approaches are widely used in cloud computing environments depending on application architecture and performance requirements. Vertical VM sizing involves increasing or decreasing the resources of a single virtual machine. This means adjusting CPU cores, RAM, storage