Energy efficiency is a key part of networking—it saves on cost and has an environmental impact. Multi-path routing is a well-known technique used for network traffic management to make it more efficient and reliable. Unfortunately, traditional multi-path routing techniques are oblivious to energy and suffer from poor energy utilization. This paper: An approach to efficient smart scheduling for multi-path routing with deep learning. Employing a deep learning algorithm with relevant network parameters allows us to predict the lightest path for processing in/outbound data. Simulation results demonstrate that our solution guarantees higher energy efficiency than conventional algorithms while enjoying better network performance. This solution proposes the formulation of a green and sustainable networking approach. Thus, it can be better adopted as an innovative security mechanism in many practical scenarios.