I, I went to front(end) and back(end)
To find myself on track
I broke the code, woah-oh-oh
— adapted from "The Code" by Nemo
Languages: Python · TypeScript · JavaScript · Bash · SQL
Backend: FastAPI · Flask · Node.js · Express · NestJS
Data: PostgreSQL · MongoDB · Redis · SQLite
AI: LLM integrations (Ollama · Hugging Face · Claude · Groq)
Infra: Linux · Docker · GitHub Actions
| Project | Description | Language |
|---|---|---|
remote-compiler |
Lightweight secure remote C/C++ compiler with end-to-end encryption for offloading builds from resource-constrained devices to powerful servers | Python |
dev-ledger |
DevLedger is a local, append-only coding activity tracking system for VS Code. It records developer events (file focus, edits, saves, and idle states) into a tamper-evident log with cryptographic hash chaining. The data is fully processed locally and can be deterministically converted into statistics and reports for README or external analysis. | TypeScript |
typecore |
Header-only C++17 library for automatic runtime type detection from strings | C++ |
adaptive-optical-modulation-ai |
This repository contains the reference implementation of the Adaptive Optical Modulation method | - |
public-appeal-service |
A REST API for managing citizen appeals to government agencies - with AI-powered classification, role-based access control, and Redis caching. | TypeScript |
ping-pong-protocol |
Lightweight LAN discovery and secure communication protocol based on a ping–pong handshake | - |
| Project | Description | Language | Stars | Forks |
|---|---|---|---|---|
ping-pong-protocol |
Lightweight LAN discovery and secure communication protocol based on a ping–pong handshake | - | ★ 2 | - |
public-appeal-service |
A REST API for managing citizen appeals to government agencies - with AI-powered classification, role-based access control, and Redis caching. | TypeScript |
★ 2 | - |
FileServer |
- | - | ★ 1 | - |
I prefer human-driven problem solving and first-principles engineering over generative code or text synthesis.
I use generative AI only where it serves as a tool for testing, validation, or integrated product functionality, not as a substitute for authorship or engineering judgment.
