offensive security · reverse engineering · systems

submerse_

Self-taught security practitioner. CompTIA Security+ Certified. I break down how systems work at the memory and protocol level — from game-engine internals and anti-cheat research to vulnerable-box exploitation, OSINT geolocation, and physical bypass work.

about

$ whoami

I learn by taking things apart. Over the last few years I've worked through a large number of vulnerable machines (top-percentile on TryHackMe), built reverse-engineering projects against game engines, and stood up my own GPU-passthrough virtualization lab to study Malware, Anti-Cheat, isolation and low-level systems.

Comfortable across Linux and Windows internals, networking, and applied cryptography — and the same investigative mindset shows up in side pursuits like OSINT geolocation and locksport. Currently Security+ certified and working towards Network+ and a practical pentesting certification (OSCP) while looking for a first role in IT/security.

capabilities

$ What I work with

Linux internalsWindows internalsReverse engineering Memory analysisx86 / assemblyC / C++ PythonNetworking / TCP-IPApplied cryptography Privilege escalationKVM / QEMUGPU passthrough OSINT / geolocationPhysical securityAnti-cheat / detection
selected work

$ Projects

↗
00

Custom Workstation Build

Planned and built from scratch — Ryzen 9 9950X3D, RTX 4090, 64Gb ddr5 6400Mhz, 360mm AIO and seven-fan airflow. The host the rest of this portfolio runs on.

hardwarethermalsassembly
↗
01

Game Internals & Anti-Cheat Research

Reverse-engineered an engine's memory model — pointer chains, entity lists, world-to-screen — to study how anti-cheat detection works.

C++REmemoryhooking
↗
02

GPU-Passthrough Virtualization Lab

Built a KVM/QEMU host with a dedicated GPU passed through to a guest — IOMMU groups, VFIO, and an isolated environment for malware/RE work.

KVMVFIOLinuxisolation
↗
03

Self-Hosted Offline LLM

A 24B model served locally on one 4090 — quantization, GPU offload, KV-cache tuning — plus red-teaming refusal robustness and where guardrails actually live.

llama.cppML-opsAI red-team
↗
04

OSINT & Geolocation

Reading locations from minimal visual cues — signage, road markings, vegetation, sun position — the same discipline used in open-source investigations.

OSINTgeolocationGeoGuessr
↗
05

Physical Security & Locksport

Single-pin picking, raking, and bypass against a range of lock mechanisms — understanding how physical access controls actually fail.

locksportphysicalbypass
↗
06

Screen-Based Detection Evasion

Why a screen-only cheat never touches game memory — and where that forces kernel anti-cheat like Vanguard to move: behaviour, server-side, hardware.

anti-cheatdetectionCV
↗
07

Detecting Hardware Input Emulation

Why an Arduino-on-a-host-shield posing as a real mouse defeats device-layer checks — and how anti-cheat like Vanguard responds: behaviour, fingerprinting, server-side.

HIDanti-cheatdetection
↗
08

CTF & Box Writeups

Methodology-focused walkthroughs of vulnerable machines — enumeration, foothold, privilege escalation, and what I'd do differently.

pentestprivescenum
contact

$ Get in touch

Email: L_s@tuta.com
PGP: public key · 5F52 F282 ECA2 F0E5 669C 1D90 2053 98B4 0315 06F3
Based in Central Switzerland · open to IT / security roles (DE / EN).