LogSentinel AI

AI-powered behavioral log analysis for intelligent threat detection.

Inspiration

Security teams are overwhelmed by massive volumes of logs, yet most detection systems still rely on static rules and known signatures. Modern attacks evolve faster than rule updates.

We built LogSentinel AI to detect malicious intent based on behavioral patterns — not predefined signatures.

What It Does

LogSentinel AI analyzes system and server log files and uses Google Gemini to:

• Detect malicious intent from behavioral summaries
  
• Classify threats (Low, Medium, High)
  
• Assign confidence scores
  
• Explain the reasoning in plain language
  
• Store results in a Django dashboard for review and tracking
  

Instead of black-box outputs, the system generates explainable threat reports.

How We Built It

• Backend: Python + Django
  
• Log Parsing: Regex and structured data processing
  
• AI Analysis: Behavioral summaries sent to Google Gemini API
  
• Database: Threat reports stored for monitoring and auditing
  
• Frontend: Django-based dashboard for visualization
  

Architecture Flow

1. Log Ingestion
   
2. Log Parsing & Structuring
   
3. Behavioral Summary Generation
   
4. AI Intent Classification
   
5. Confidence Scoring
   
6. Dashboard Visualization
   

Example Log Input

Failed login attempt from 192.168.1.10
Multiple access requests within 3 seconds
Admin privilege escalation attempt

Built With

• gemini
• python