We built S.I.C.R.O. (Intelli-Net) as a minor project earlier this year — a system that pulls network telemetry straight from Cisco DNAC's REST API and puts it in front of us through a Streamlit dashboard, backed by SQLite for storage, Groq for the intelligence layer, and Discord for alerts. At the time, the goal was simple: get visibility into what's happening on the network without digging through raw logs.
Now that we're heading toward a security-focused final year project, the natural next step is obvious — we're already pulling live network data. What if instead of just displaying it, the system started asking "does this look wrong?"
Here's the plan for how we want to extend it:
1. Add anomaly detection on top of existing telemetry
Instead of just showing metrics, we want to baseline "normal" traffic patterns and flag deviations — unexpected spikes, unusual device behavior, or new devices appearing on the network. Nothing fancy at first, just threshold and statistical-based detection before we consider anything ML-driven.
2. Turn Discord alerts into SOC-style alerting
The alert pipeline already exists for basic notifications. We want to restructure it to mimic how a real SOC prioritizes alerts — severity tagging, so it's not just "something happened" but "something happened, here's how bad it might be."
3. Add a basic incident log
Right now the dashboard shows live state. We want to add a persistent incident history — what got flagged, when, and whether it was resolved or a false positive. This is closer to how ticketing works in an actual SOC workflow.
4. Map features to real SOC concepts
The bigger goal isn't just "add security features" — it's to consciously mirror what SOC analysts actually do: monitor, detect, triage, escalate. Each feature we add should map to one of those steps so the project reads as a genuine security tool, not just a dashboard with extra bells on it.
We're documenting this as we go, partly because writing it down forces us to actually think through the plan instead of vibing our way to the final review. If anyone's extended a networking/monitoring project into a security one before, would love to hear what worked and what didn't.
Loading article