Sarcouncil Journal of Engineering and Computer Sciences

Abstract: AI-driven sales forecasting has emerged as a transformative capability within modern CRM systems, enabling more accurate, adaptive, and explainable predictions for revenue and deal outcomes. This review investigates how machine learning, pipeline intelligence, and predictive modelling are integrated into platforms like Microsoft Dynamics 365 Sales. Drawing on a decade of academic and industry research, it presents a humanised summary of methodologies, results, and real-world use cases, showcasing tangible improvements in forecast accuracy, deal prioritisation, and sales planning. The review also introduces a theoretical model for predictive pipeline forecasting and outlines emerging trends, including generative AI, online learning, and responsible AI governance. Together, these advancements suggest a shift toward real-time, trustworthy, and hyper-personalised sales forecasting ecosystems. In fact, industry studies report that companies adopting AI-driven forecasting achieve nearly 79% accuracy on average versus roughly 51% for those relying on traditional methods. This highlights how AI and pipeline intelligence are elevating forecast reliability in practice.