SentinelAI – AI Cybersecurity Threat Detector

SentinelAI is an AI-powered cybersecurity platform built to automate the analysis of security logs and accelerate threat detection for security analysts. Organizations generate thousands of log entries every day from authentication systems, web servers, and firewalls, and manually reviewing them is slow, inconsistent, and easy to get wrong. SentinelAI solves this by letting users upload common log formats — auth.log, syslog, apache.log, firewall.log, CSV, and JSON — directly through a web dashboard. Once uploaded, the FastAPI backend parses the log file and passes it to an AI analysis engine powered by Fireworks AI (Llama 3.1 70B Instruct). The AI identifies potential threats, assigns a severity level (Low, Medium, High, or Critical), calculates an overall risk score, and maps detected attack patterns to the MITRE ATT&CK framework. Results are stored in a PostgreSQL database and surfaced through a real-time dashboard showing severity distribution, top MITRE techniques, and recent incidents. Beyond detection, SentinelAI automatically generates structured incident reports summarizing findings and recommended next steps, and keeps a searchable history of every past analysis. An integrated AI chat assistant lets analysts ask natural-language questions about threats, request explanations of attack techniques, or get prevention guidance, turning raw detection output into actionable security intelligence. Built with a modern Next.js/React frontend, a FastAPI Python backend, JWT authentication, and Docker-based deployment, SentinelAI demonstrates how AI can meaningfully reduce manual security operations workload while giving analysts clearer, faster insight into what matters most in their logs.