EfficientML-Cloud: Hybrid Cost-Aware Inference for Enterprise ML Systems

EfficientML‑Cloud introduces a hybrid cost‑aware inference framework designed for enterprise machine‑learning systems deployed across cloud and on‑premise environments. The approach dynamically routes inference requests between local compute and cloud services based on latency, cost, and workload conditions. By combining lightweight local models with scalable cloud inference, the system reduces operational expenses while maintaining high accuracy and responsiveness. The work includes architectural design, routing algorithms, and evaluation results demonstrating improved efficiency under real‑world workloads.