AI-Driven Security and Inventory Optimization: Automating Vulnerability Management and Demand Forecasting in CI/CD-Powered Retail Systems

The work focuses on examining how artificial intelligence (AI) can be applied to solve two of these key issues in contemporary retail concurrently, that is, vulnerability management automation in continuous integration/ continuous delivery (CI/CD) pipelines and inventory optimization based on demand forecasting. Retail organisations have been turning more and more to CI/CD to facilitate fast delivery of features and security upgrades. The acceleration causes an expanded attack surface, making vulnerability management harder. Meanwhile, considering the demand error forecasting causes expensive stockouts or excessive stock. The study uses a dual-framework approach, which utilizes heterogeneous datasets such as point-of-sale (POS) transactions, enterprise resource planning (ERP) data, and vulnerability feeds that comprise the National Vulnerability Database (NVD) and Common Vulnerabilities and Exposures (CVE). Sophisticated models that include Random Forests, Support Vector Machine, and Long Short-Term Memory (LSTM) are used to predict and classify vulnerabilities and enhance demand forecasts beyond those of conventional statistical models. The experimental validation proves the ability of AI-driven triaging to decrease the patching delays and the mean time to remediation (MTTR), and of deep learning to increase the accuracy of the forecasting, leading to increased inventory availability. Future directions identified in the study include reinforcement learning to schedule patches to allow adjustable scheduling, edge AI and IoT-driven real-time forecasting to allow a just-in-time replenishment, and immutable logging via blockchain to enable secure vulnerability management and supply chain traceability. All findings together prove the idea that AI can support the resilience of cybersecurity, as well as effectiveness in CI/CD-driven retailing ecosystems.