skip to content →
system online — hunair.fun

Hunair_

$whoami

Welcome to my digital playground. Securing the web and teaching machines new tricks — one commit, one prompt, one exploit at a time.

0+
commits pushed
0+
bugs squashed
0
models trained
0
cups of coffee

// system notice

The Lab is under construction

experiments are being calibrated. check back soon for the full skills manifest.

coming soon

// ./projects --featured

Featured Projects is under construction

I'm rebuilding this section with case studies, metrics, and live demos. Check back soon.

coming soon

// ./access --shell

Live Shell

poke around. type help for commands. arrow keys scroll history. tab autocompletes.

hunair@hunair.fun: ~/shell

 _   _                  _ _ _
| | | | ___ _   _  __ _| | | | ___
| |_| |/ _ \ | | |/ _' | | | |/ _ \
|  _  |  __/ |_| | (_| | | | |  __/
|_| |_|\___|\__, |\__,_|_|_|_|\___|
            |___/   :: hunair.fun v2.0
type 'help' to list commands.
visitor@hunair.fun:~$
↑↓ history · tab complete · ctrl+l clear

// ./timeline --reverse

The Journey

  • 2019

    first compile

    Wrote my first Python script — a port scanner that crashed my router. Hooked ever since. Spent more time in `/var/log` than in class.

  • 2021

    red team awakening

    Joined a CTF on a whim, captured a flag in the crypto category at 3am, and never looked back. Began formal pentesting work shortly after.

  • 2023

    machine whisperer

    Trained my first transformer from scratch on a single GPU. Realized AI security was the wild west — and that's exactly where I wanted to be.

  • now

    building in public

    Shipping tools at the intersection of security and intelligence. Breaking things to understand them, then rebuilding them better, faster, safer.

[ H ]

// whoami

About Me

I'm Hunair — a developer working at the seam between security and artificial intelligence. I break things to understand how they work, then build intelligent systems to put them back together better. Some days that means dissecting malware; other days it means training a model to write better exploits than I can. Both days are good days.

My happy place is the overlap: adversarial ML, prompt-injection defense, autonomous red-team agents, and the occasional side-quest into creative visualization. I believe the next decade of security will be fought with models — and I'd rather be on the team building the shields than the one mining the exploits.

based in
earth // remote-first
stack
python · ts · rust · go
focus
ai × security
fuel
caffeine + curiosity
PENETRATION TESTINGLLM RED-TEAMINGAGENTIC SYSTEMSREVERSE ENGINEERINGZERO-KNOWLEDGE CRYPTONETWORK DEFENSEPROMPT ENGINEERINGWEBGL VISUALIZATIONSCTF PLAYEROPEN SOURCEPENETRATION TESTINGLLM RED-TEAMINGAGENTIC SYSTEMSREVERSE ENGINEERINGZERO-KNOWLEDGE CRYPTONETWORK DEFENSEPROMPT ENGINEERINGWEBGL VISUALIZATIONSCTF PLAYEROPEN SOURCE