I work on low-level systems, firmware reverse engineering, AI/NPU experimentation, security research, and developer tooling.
I like projects that answer a concrete technical question with reproducible code, documented evidence, and clear limitations.
- Firmware reverse engineering and embedded systems
- AMD XDNA / Ryzen AI NPU experimentation
- Low-level Linux and hardware research
- Security analysis and reverse engineering
- AI-assisted developer tooling and automation
Reproducible reverse engineering of Samsung Odyssey G5 G55C firmware, including architecture mapping, validation tooling, safety analysis, and a minimal data-only factory-page unlock.
Experiments with LLM workloads across AMD XDNA1 NPU, CPU, and GPU on Ryzen AI Linux systems.
Low-level experiments around compute workloads on AMD XDNA NPU hardware.
Evidence-based static security analysis with hashes, disassembly, indicators of compromise, and YARA rules.
- Reproducible results over vague claims
- Evidence before conclusions
- Small, auditable changes
- Clear documentation of limitations
- Automation where it removes repetitive work
Browse my public repositories at @c8dhjp4tyv-bit.
Research, break things carefully, document what actually happened.


