Development, IT support, networking & security

Jason S. Sec. & Sys.

I set up, break, and fix real systems in my homelab, from DNS filtering and remote access to a small routed network. This is where I show what I built and what I learned along the way.

EVE-NG / AREA 0 INFO

Available for junior roles

6:57 AM EDT · New Jersey

The window follows New Jersey sunrise, sunset and weather. The laptop screen is a live display of my lab network.

cd ~/lab/rack

Work from the lab.

01 DNS DETECTOR
Detection

DNS Exfiltration Detector

I found that a public DNS exfiltration benchmark's attack data contains no attacks, rebuilt it from the raw packet captures, and showed where a model trained on it breaks.

CASE STUDY →
CAUGHT, DISGUISED EXFIL2.2%SAME MODEL: 100% ON BENCHMARK
Benchmark attacks · ~60-char base64 labels100%
Synthetic exfil · 8-char hex labels2.2%

The model learned “long and random,” not “exfiltration.”

02 NETOPS AGENT
Network automation

Homelab NetOps Agent

A small local model (Qwen2.5-3B) that answers questions about my OSPF lab by running read-only commands on the real routers and quoting what comes back. No agent framework, just Python and SSH.

CASE STUDY →
TASKS PASSED / 127.1→7.9BEFORE → AFTER FINE-TUNING
› How many OSPF neighbors does R1 have?
$ show ip ospf neighbor R1
Neighbor ID State Up Time Address
10.50.0.2 Full/DR 4d21h37m 10.0.12.2
3.3.3.3 Full/DR 4d21h37m 10.0.13.2
R1 has 2 OSPF neighbors.
GRADED: PASSED · RUN 2026-08-18
03 HOME NETWORK
Home network

Home Network Hardening

I set up Pi-hole and Tailscale so my phone uses the same DNS filter at home and away. I also used ADB to clean up a carrier-issued Android TV box.

CASE STUDY →
TARGET TRACKERS BLOCKED4/4ON ANY NETWORK
PI-HOLE · TAILSCALE · ADB
Targeted tracker domains4/4 BLOCKED
Phone on cellular and outside Wi-FiPI-HOLE DNS
Inbound router ports opened0
Old TV launcher restarted byUID 0 · DISABLED
04 LIF SNAKE
Spiking neurons

LIF Snake

A Snake controller built from leaky integrate-and-fire neurons and synapses that learn from reward. No ML libraries. Every spike and weight update is my own code.

CASE STUDY →
FOOD PER GAME19.52UNTRAINED: 0 · 40 UNSEEN SEEDS
RECORDED DECISION · STEP 001REWARD +0.08
heading eastfood westfood southpath clear
LEFT
STRAIGHT
RIGHT

Turns right: the motor with the most spikes wins.

05 LIVE LAB
Hybrid cloud

Live Lab Pipeline

My OSPF lab at home reports through a Google Cloud watchdog to this site. Router passwords never leave the house.

CODE ON GITHUB →
ADJACENCIES3/3CHECKING THE LAB…
R1–R3 · EVE-NGnode1GCP watchdogjason-ss.dev
LAB LOG

Waiting for the first report.

+ ALL PROJECTS8 PROJECTS →

ls ~/projects

Everything I've built.

open resume.pdf

Resume.

My current résumé, with experience in data analysis, reporting, networking, and operations support.

whoami

I like finding out why things break, and then fixing them.

I graduated from Rutgers–Newark and NJIT in May 2026 with a B.A. in computer science. Most of my hands-on IT experience comes from my homelab: two Ubuntu mini PCs, Pi-hole for DNS filtering, Tailscale for remote access, and a three-router network lab in EVE-NG. When something stops working, I check the logs, change one thing at a time, and write down what fixed it.

I'm also careful with my own results. On my DNS exfiltration project, I found that a public benchmark's attack data contained no attacks, so I rebuilt it from the raw packet captures before trusting any score. I'm looking for an entry-level role in IT support, help desk, or junior network, security, or GRC work, on a team where I can learn from people with more experience. I'm studying for CompTIA A+, I speak English and Spanish, and waiting tables taught me how to stay patient and clear with someone who just needs their problem fixed.