Elevate: A Unified Web-based Platform for Ransomware Detection and Network Intrusion Analysis

Abstract

The rapid evolution of cyber threats, including advanced ransomware, zero-day exploits, and sophisticated network intrusions, has exposed the limitations of traditional security systems that rely primarily on detection without real-time response capabilities. This study presents a comprehensive survey and design perspective for a unified AI-driven cybersecurity platform that integrates machine learning, network monitoring, and autonomous response mechanisms to enhance system resilience. The proposed framework is designed using a modular architecture that combines machine learning-based intrusion detection systems, ransomware detection engines, and real-time network traffic analysis. It incorporates datasets such as UNSW-NB15 and leverages algorithms like XGBoost to identify anomalous patterns in system behaviour and network flows. Additionally, the system integrates local large language models (LLMs) via Ollama to enable intelligent threat interpretation, automated log analysis, and AI-assisted ransom negotiation. Furthermore, the platform includes automated backup and recovery mechanisms, enabling self-healing capabilities in the event of an attack. Behavioural analysis, forensic log monitoring, and proactive defence strategies are employed to detect and mitigate threats before significant damage occurs. The system also emphasizes privacy preservation by performing critical computations locally, reducing reliance on cloud-based processing.