Systems Researcher | Compilers & Distributed Runtimes
"The more you know, the more you know you don't know."
Note
This is a personal profile README — a running log of things I’ve built, broken, and learned.
Not a professional portfolio. Not here to impress anyone.
Graduating June 2026 (BSIT) — Colegio de Montalban
Formal Methods • Runtime Security • Systems Architecture
I enjoy building systems from scratch to understand how they work under the hood. This profile is just a collection of my projects and experiments—ranging from custom runtimes to security research.
- 🎓 Current Status: Graduating June 2026 (BSIT) | Colegio de Montalban.
- 🦀 Focus: Formal Methods, Runtime Security, Systems Architecture.
- 🧪 Primary Stack: Rust (Performance Primitives), Zig (Native Interception), Python & JavaScript (Orchestration).
- 🌐 Web Development: Engineering modern, performant web applications utilizing Next.js and FastAPI.
- 🫢 Notice: I prefer building something from scratch over reading about it. Currently diving deeper into formal methods and runtime internals.
A collection of theoretical and practical explorations into systems-level engineering.
| Repository | Project Specification | Technologies |
|---|---|---|
| Lila | A formally verified, affine-typed JIT compiler for Python. By leveraging the Z3 SMT solver and Cranelift, Lila enforces Liquid Types (logical invariants) and Fractional Permissions (memory safety) at compile-time, enabling GIL-less execution of Python at bare-metal speeds. | |
| Astraea | A zero-trust security middleware for Node.js implementing an Object-Capability (O-Cap) enforcement layer. It utilizes dynamic linker hijacking via a Zig interceptor and a Rust engine to perform context-aware attribution and Seccomp-BPF kernel hardening against supply-chain attacks. | |
| Iris | A high-performance distributed actor runtime in Rust with first-class Python bindings. It features a reduction-based cooperative scheduler (Vortex) and a location-transparent messaging fabric, enabling zero-downtime hot-swapping of actor logic and resilient, cross-language service orchestration. | |
| Nimue | A selective Rust virtualization framework (co-authored) utilizing a polymorphic Mother/Worker VM architecture and the Prism non-linear string protection system for advanced software obfuscation. | |
| Causm | A domain-specific research language targeting non-determinism in concurrent systems. It implements an Entropic Memory Model (state decay) and Isochronous Scheduling, using a Z3-governed correctness kernel to mathematically prove temporal invariants and race-freedom at compile-time. | |
| Ocular | A high-performance, zero-allocation instruction tracer for Python 3.12+ (PEP 669). It leverages hardware TSC cycle counting for nanosecond-precision profiling and implements dynamic de-instrumentation, automatically unhooking from hot loops to allow CPython's native quickening while generating Chrome Perfetto timelines. | |
| MirageFS | A steganographic virtual block device enabling encrypted storage within multimedia containers. By camouflaging data as Adobe DNG metadata or H.264 NAL-unit filler, it creates a FUSE-mounted filesystem that is mathematically indistinguishable from standard media noise to defeat forensic analysis. | |
| Argus | An entropy-based secret scanner utilizing Aho-Corasick pattern matching and Shannon entropy context to distinguish true cryptographic risks from noise in large codebases. | |
| Project Aether | A userspace TCP/IP stack implementation in Rust designed to bypass kernel overhead, prioritizing raw throughput for specialized, high-performance networking research. | |
| Sigilus | A multi-stage WebAssembly-based security framework for web APIs, implementing cryptographic proof-of-execution challenges to neutralize automated threat vectors. |
I'm a big believer in learning by building. To really understand a complex concept, I try to implement it from scratch.
- Systems & Runtimes: Rust is my go-to for memory safety and performance. I use it to build core infrastructure like distributed actor runtimes and custom schedulers, pairing it with Python and JavaScript for high-level orchestration.
- Compilers & Formal Methods: I spend a lot of time exploring language design and compiler architecture. I'm especially interested in integrating SMT solvers (Z3) to mathematically prove memory safety, type constraints, and temporal invariants.
- Security & Low-Level Internals: I like digging into the OS boundary. This involves native C-ABI interception (often using Zig), kernel-level sandboxing (Seccomp-BPF), and researching steganography and anti-forensics.
- Web Architectures: When building standard applications or APIs, I stick to reliable stacks like Next.js and FastAPI, focusing on type safety and fast development cycles.
- Workflow: Mostly terminal-based. I use KDE Kate for most of my editing, but switch to VSCode when I'm working on language support or larger projects.
I'm always down to chat about systems, performance (like why Python is slow), or nerdy/weeb stuff. If you're building something weird or unconventional, feel free to reach out.




