Open to new roles

Systems Architecture Engineer

LEELANDRAJKUMAR

I build internal platforms that teams actually use

TypeScript · React · Node · Python · PostgreSQL · Linux · nginx · systemd · Docker

Full-Stack Developer • Johor Bahru, Malaysia

Leeland Raj Kumar
Scroll

Get to know me

About Me

I'm a Systems Architecture Engineer based in Johor Bahru, Malaysia. I build internal platforms, the kind that replace messy, fragmented processes with something reliable that people actually want to use.

At Wiwynn, I took a platform from idea to company-wide deployment mostly on my own. That meant owning everything: system design, database schema, REST API, frontend, access control, and keeping it all working as more teams picked it up.

Outside of work I run my own server, and it hosts almost everything I build, not local demos sitting on a laptop. Being the one who has to keep those things patched, backed up, and running at 3am changes how you write the code in the first place. Shipping something is one lesson; staying responsible for it afterward is a different one.

Currently looking for my next role in systems engineering, full-stack development, or platform work. Open to Malaysia and international opportunities.

1

Platform deployed company-wide at Wiwynn

7

Meta Professional Certifications

25

services I run in production

Let's collaborate to build something truly innovative!

Get in touch →

The thread

My Journey

Where it started

Systems and low-level work — C++, assembly, computer architecture. The kind of code where correctness isn't optional. That foundation shaped how I think: structure first, implementation second.

What I built

At Wiwynn, I took fragmented manual workflows and replaced them with a single internal platform — end to end. System design, database schema, REST API, frontend, access control, deployment. Solo. Company-wide adoption. Outside that job I've kept building on my own terms: real-time systems, local ML and LLM integration, and a server I run and maintain myself so I can see how the things I build actually hold up once something depends on them.

What I'm looking for

My next role in systems engineering, full-stack development, or platform work. Somewhere the architecture decisions actually matter and the team takes ownership of what they ship. Open to Malaysia and remote opportunities.

Technical expertise

Skills

Frontend

React / TypeScriptNext.jsTailwind CSSThree.js / R3FTanStack Query

Backend & Data

Node.js / ExpressPython / FastAPIPostgreSQLSQLiteREST API DesignJWT / Auth

Infrastructure

LinuxnginxsystemdDockeroauth2-proxy SSOTLS / certbot

Real-time & AI

Server-Sent EventsSocket.IOWebSocketsLocal LLM (Ollama)CLIP / onnxruntimeClaude API

My work

Projects

← Scroll or drag to explore →

Self-hosted

The Lab

This is a personal Ubuntu server I administer end-to-end — my own machine, not employer infrastructure. Everything on it is operated the same way I'd operate a production system: systemd supervision, TLS everywhere, SSO in front of anything private, and automated backups.

Personal homelab — request flow

Homelab request-flow architecture diagramTraffic flows from the internet through nginx, which terminates TLS for 16 subdomains. Four of those subdomains sit behind an oauth2-proxy single sign-on layer before reaching systemd-managed Node.js and Python services, which persist to SQLite and PostgreSQL. Two side branches hang off the services layer: a self-hosted Pelican and Wings panel orchestrating nine game servers, and a local Ollama LLM used by two in-house tools.InternetnginxTLS termination · 16 subdomainssystemd servicesNode.js + Python · 25 unitsSQLite + PostgreSQL8 SQLite DBs · nightly backupsoauth2-proxySSO layer · 4 of 16 subdomainsPelican + Wingsgame panel · nightly cron restartsOllama (local LLM)qwen2.5:1.5b-instruct2 in-house tools · on-box only9 game servers7 Minecraft · 1 Palworld1 Satisfactory
25
self-managed services in production
0 crash-triggered restarts · zero failed units
4wk 1d
uptime at last measurement
longest-running services unbroken since Aug 21, 2026
16
TLS-secured subdomains
11 auto-renewing Let’s Encrypt certificates · 4 behind SSO
840,242
external HTTP requests served
15-day window (Sep 5–19, 2026) · 485,488 in the most recent 7 days · ~3,949 unique client IPs
9
game server instances orchestrated
7 Minecraft · 1 Palworld · 1 Satisfactory, via self-hosted Pelican + Wings
8
live SQLite databases
nightly backup rotation on the two highest-traffic apps

How it's run

  • systemd units with restart policies — zero crash-triggered restarts across all 25 services
  • A certbot systemd timer auto-renews all 11 Let’s Encrypt certificates before they expire
  • oauth2-proxy SSO gates every subdomain that isn’t meant to be public
  • Nightly backup rotation on the two highest-traffic SQLite databases
  • Custom systemd timers: a metrics collector every ~15–20s, a Minecraft TPS health monitor every 5 min

Credentials

Certifications

Meta Back-End Developer Specialization

Meta

Meta Back-End Developer Specialization

The Full Stack

Meta

The Full Stack

Version Control

Meta

Version Control

APIs

Meta

APIs

Programming in Python

Meta

Programming in Python

Back-End Developer Capstone

Meta

Back-End Developer Capstone

Django Web Framework

Meta

Django Web Framework

Let's talk

Contact Me

Let'sbuildsomething.

Have something in mind? I'm open to collaborations and full-time opportunities.

Email

Get in touch for collaborations or inquiries. I typically respond within 24 hours.