Modern computing systems require efficient power management since power consumption directly affects battery life, performance, and temperature handling. Dynamic Voltage and Frequency Scaling (DVFS) is a popular technique that uses dynamic processor voltage and frequency scaling to balance power and performance. In order to optimise DVFS scaling decisions, this study looks into a machine learning-based approach that prioritises power-conscious frequency prediction while maintaining system performance. Traditional DVFS methods typically rely on preset thresholds or static rules, which may not be adaptable enough to shifting workloads or dynamic system conditions. In contrast, the AI-powered DVFS model learns from system behaviour and historical experiences to make more intelligent, context-sensitive modifications. We developed an interactive HTML web page that shows performance and power usage measurements in order to illustrate the effectiveness of the proposed solution. According to the results, the AI-based method is more responsive and energy efficient than conventional DVFS techniques. It is also a promising advancement for low-power computing systems of the future.