██╗ ██╗ ██████╗ ██╗ ██╗██╗███████╗ ███████╗██╗ ██╗███╗ ██╗
██║ ██║██╔═══██╗██║ ██║██║██╔════╝ ██╔════╝██║ ██║████╗ ██║
███████║██║ ██║██║ █╗ ██║██║█████╗ ███████╗██║ ██║██╔██╗ ██║
██╔══██║██║ ██║██║███╗██║██║██╔══╝ ╚════██║██║ ██║██║╚██╗██║
██║ ██║╚██████╔╝╚███╔███╔╝██║███████╗ ███████║╚██████╔╝██║ ╚████║
╚═╝ ╚═╝ ╚═════╝ ╚══╝╚══╝ ╚═╝╚══════╝ ╚══════╝ ╚═════╝ ╚═╝ ╚═══╝
Chief Technology Officer @ Kinepara
Kernel Security · Adversarial Systems · Algorithmic Engineering
I approach systems engineering as the disciplined intersection of algorithmic rigor, kernel-level security, and adversarial resilience — where correctness is a prerequisite, not an afterthought.
- Implementing wait-free and lock-free primitives (Ring Buffers, Skip Lists, Hazard Pointers) under strict atomic memory ordering
- Mitigating ABA problems, eliminating false sharing, and optimizing working-set locality via van Emde Boas layouts and CSR compression
- Reducing amortized complexity from
$\mathcal{O}(n \log n)$ to$\mathcal{O}(n)$ through radix-based decomposition and cache-oblivious traversal strategies
- Researching privilege escalation vectors: TOCTOU races, use-after-free in kernel object lifecycles, and improper capability / namespace isolation
- Hardening kernel subsystems via seccomp-BPF policy design, LSM hook instrumentation, and mitigating speculative execution side-channels (Spectre/Meltdown variants)
- Analyzing syscall interception and DKOM (Direct Kernel Object Manipulation) as both attack surfaces and detection primitives
- Designing kernel-mode integrity monitors leveraging PatchGuard-equivalent callback chains and SSDT hook detection
- Applying Bloom Filters and Count-Min Sketches for high-throughput behavioral stream analysis under strict latency budgets
- Modeling cheat detection as an adversarial classification problem — balancing false positive rate against evasion resistance under a game-theoretic threat model
$\mathcal{G} = \langle \mathcal{A}, \mathcal{D}, \mathcal{U} \rangle$ - Leveraging hypervisor-assisted introspection (VMI) for tamper-evident memory scanning beneath the OS trust boundary
- Optimizing flow, matching, and shortest-path algorithms for latency-critical detection pipelines
- Applying DP over DAGs and number-theoretic primitives to integrity verification and signature schemes
| Competition | Distinction | Domain |
|---|---|---|
| National (NOI) | Medalist (Top Tier) | Advanced Data Structures & Algorithms |
| Collegiate (ICPC/CCPC) | Regional Medalist | Combinatorics & Optimization |
| Provincial | Multiple First Prizes | Heuristic Search & Number Theory |
Reliability is not an accident.
fn architecture() -> Result<System, Entropy> {
match design {
Design::FirstPrinciples => Ok(Scalability),
Design::AdHoc => Err(TechnicalDebt),
}
}Type systems and formal reasoning over runtime checks. Correctness by construction.
Deep technical discourse on:
├── Systems Programming
├── Kernel Security Research
└── Adversarial / Anti-Cheat Engineering
"Talk is cheap. Show me the code." — Linus Torvalds


