PROJECT NOTES / 03
TriDetect
Structured logs from the web, application and database layers, brought together in a Python detection pipeline and a Streamlit dashboard.
The question behind the project
Looking at one application layer can hide the rest of an incident. TriDetect brings signals from a web tier, an application tier and a database into the same monitoring workflow.
What I built
- Used Docker Compose to organize the three tiers and a shared structured-log volume.
- Built around a Python polling agent with pattern matching and thresholds for SQL injection, XSS, repeated authentication failures and automated scans.
- Presented the aggregated alerts in Streamlit, with views for event severity and cross-tier activity. The repository includes the agent, dashboard and controlled attack-simulation script.
What came out of it
The result is an explorable monitoring application with source code and original dashboard captures. The project links application development with log analysis rather than treating security as a separate interface.
What the evidence does—and doesn’t—say
The demonstrations use a controlled environment and synthetic attacks. They are not a production detection benchmark.
From the project
Select a capture to inspect it.
07 / CONTACT
Let’s start with
a conversation.
Have a PFE internship in cybersecurity, networks or infrastructure security? I’d be glad to hear about the work and the team.
Marrakech, Morocco