HawkSentinel - Autonomous Security Watchdog

HawkSentinel AI is an autonomous, real-time account security system designed to protect large-scale trading platforms like Deriv from account takeovers and unauthorized access. Modern attacks often bypass static security rules by mimicking normal user behavior. HawkSentinel addresses this gap by continuously analyzing live login events using device consistency, geo-velocity, and behavioral patterns. Each session is risk-scored in milliseconds, allowing the system to identify anomalies such as impossible travel or previously unseen high-risk device configurations. What differentiates HawkSentinel is its AI Forensic Agent, which does not simply raise alerts. Instead, it explains why a session is risky in plain language and recommends immediate actions. For high-risk events, HawkSentinel can autonomously freeze accounts and trigger two-factor authentication resets without waiting for manual review. The system is built using Python, FastAPI, Streamlit, and large language models, and runs entirely on synthetic data for safe demonstration. HawkSentinel demonstrates how explainable AI can enable faster, safer, and more trustworthy security decisions at scale.