CloudPilot: AI-Based Cloud Advisory, Cost Optimization, Usage Forecasting, and Beginner-Centric Resource Management

Abstract

Cloud computing has become the backbone of modern software development, academic research, and machine learning experimentation due to its scalability and on demand resource availability. However, cloud adoption introduces significant challenges for beginners such as students, researchers, and early-stage developers, primarily due to complex pricing models, improper service selection, and lack of expert guidance. These challenges often result in inefficient resource utilization and unexpected billing overheads. This paper presents CloudPilot, an AI-based beginner-centric cloud advisory and real-time cost optimization platform designed to bridge the knowledge gap between cloud infrastructure complexity and user requirements. CloudPilot assists users in selecting appropriate cloud services based on project needs and budget constraints, generates cost-optimized cloud architectures with step-by-step deployment guidance, and continuously monitors real-time cloud usage and billing data after deployment. The system employs time-series forecasting and anomaly detection techniques to predict future cloud expenses, detect inefficient resource utilization, and recommend optimization actions proactively. Experimental evaluation demonstrated that CloudPilot achieved 93.2% cloud cost forecasting accuracy and reduced unnecessary cloud expenditure by 31% through proactive monitoring and optimization recommendations. The proposed platform improves accessibility and affordability for beginner cloud users while enabling efficient resource utilization.