WatchDog
Powered by Agentic AI

Your network never
sleeps. Neither does
WatchDog.

Autonomous AI agents that detect, classify, prioritize, and report threats in real time — before your team even opens their laptop.

Get Started → Log In
4
AI Agents
< 2s
Detection Latency
Real-time
Threat Reporting
MITRE
ATT&CK Aligned
About WatchDog

Agentic AI that thinks like a senior analyst

WatchDog is a next-generation SIEM platform built on a multi-agent AI pipeline. Each alert is automatically fetched, classified by a trained ML model, prioritized by behavioral indicators, and documented with a full incident report — all without human intervention.

Multi-Agent Pipeline
Fetcher → Classifier → Prioritizer → Reporter — four specialized agents working in concert.
Real-Time Detection
Continuous polling of Wazuh alerts with sub-2-second classification latency.
ML Classification
TF-IDF vectorized model trained on real alert data with tunable confidence thresholds.
LLM Incident Reports
GPT-4o-mini generates professional SOC incident reports for every true positive automatically.
The Team

Built by security-minded engineers

Supervisor
Dr. Rayan Mosli
Dr. Rayan Mosli
Project Supervisor
Provided the team with deep expertise in agentic AI architectures and their application in cybersecurity. Guided the integration of LLM-driven threat analysis with real-world SOC workflows, shaping the multi-agent design that powers WatchDog. His mentorship bridged cutting-edge AI research with practical security engineering.
Mohammed
Mohammed
Lead Engineer & AI Architect
Designed and built the multi-agent pipeline and the ML classification engine. Passionate about applying agentic AI to real-world cybersecurity challenges.
Nawaf
Nawaf
Security Analyst & Backend
Expert in SIEM operations and threat hunting. Led the Wazuh integration and designed the prioritization logic using MITRE ATT&CK behavioral indicators.
Mustafa
Mustafa
Frontend & Dashboard Engineer
Built the real-time dashboard and data visualization layer. Focused on making complex security data accessible and actionable for SOC analysts.