ZeroDayPulse – AI-Based Real-Time CVE Prediction System

ZeroDayPulse is an AI-driven framework that aims to accelerate vulnerability triage by combining live CVE feeds, lightweight natural language processing (TF-IDF), and ensemble classification (Random Forest). The system automatically ingests CVE records from the National Vulnerability Database (NVD), extracts textual and categorical features (description, attack vector, privileges, etc.), and computes probabilistic severity estimates. Predictions are delivered through an interactive dashboard to help security teams prioritize patching and monitoring workflows. Using automated feature extraction and an ensemble classifier produces fast, explainable scores suitable for continuous integration into vulnerability management processes. The NVD is the canonical source of CVE data used in this work; its structured feeds make automated ingestion and reproducible experiments possible.