Software Engineer | AI Researcher | Systems Optimizer

Hi, I'm Quan. I am a software developer and a deeply passionate AI researcher.

My journey into software engineering wasn't paved with high-end servers, massive compute clusters, or expensive gear. Growing up in a family without a high income, my only window into the world of technology for years was a struggling 2016 laptop with just 2 CPU cores, 8GB of RAM, and absolutely no GPU.

I experienced firsthand the frustration of hardware limitations—the endless waiting, the crashes, the inability to run modern frameworks. But that struggle became my greatest teacher. When your resources are practically zero, you don't rely on bloated abstractions; you learn to dig deep into the bare metal. You learn to squeeze every single drop of performance out of the hardware you have.

That is the soul behind my brand, Pomai (Pomegranate). A pomegranate is made of many tiny, fragile seeds. Individually, they are small and vulnerable. But when tightly bound together by a protective membrane, they form a complete, resilient, and beautiful fruit.

This GitHub profile is the living diary of my work. Everything I build here is driven by a single, unwavering mission: to build profoundly optimized, high-performance tools for individuals like me. For the students, the dreamers, and the developers who don't have money or powerful hardware, but who still deserve to run bleeding-edge software at absolute maximum speed.

I don't just use existing frameworks; I build core technologies from the ground up to ensure they can run on almost anything. Here is the ecosystem I am currently developing:

High-performance, Low-resource Database Engine. Designed for robust data storage and retrieval on constrained hardware. Built with a fierce focus on custom internal data structures, hyper-optimized indexing, and minimizing disk I/O to keep older machines breathing easily while processing heavy data.

Custom Memory Allocator. When you only have 8GB of RAM, every byte is sacred. palloc is a highly optimized memory management library designed to replace standard malloc/free. It ruthlessly cuts down memory fragmentation and allocation overhead for performance-critical applications.

Core processing engine & logic system. A specialized, lightweight system integrating internal submodules (including palloc and context-squeezing mechanisms) to handle complex AI logic and operations without melting your CPU.

• Systems Programming: C++, Assembly, Memory Management, OS Development
• AI & Machine Learning: Deep Learning on constrained devices, Data Mining, Computer Vision
• Architecture: System Design, Multi-threading, Low-latency computing, Hardware-sympathetic coding
• Infrastructure: Database Internals, Custom Allocators, Bare-metal optimizations
• Tools: Git, Linux internals, QEMU, GDB, CMake