KEVIN TAN
System Initialization
0%

System Initialization

The Architect
& The Artifact
Are The Same.

The website itself is the proof of the programmer’s thinking. Interaction, layout, and motion communicate philosophy: precision, systems thinking, and deliberate design decisions.

Kevin Tan Hong Ann

Software Developer & Technical Leader

Petaling Jaya, Malaysia (3.1073° N, 101.6067° E)

GitHubLinkedInEmail

01. Work

Case Structures

Sys.001

AI-Driven Tech Architecture

Technology partner for Mole and Reka Studios, architecting event organizing products, real estate dashboards, and virtual try-on platforms.

StackSpring Boot, Vue.js, React, PostgreSQL, Python, AWS, AI Tools
ConstraintRapid parallel development across multiple distinct industry domains requiring high context switching.
ResultIncreased productivity and delivered robust solutions using AI and vibe engineering techniques.
Architecture Diagram[Fig. 1]
AI Coding Tools
→
Microservices
↓
Multi-Domain DB Cluster

Sys.002

Print Industry Quotation Engine

SaaS quotation platform built from scratch for the printing industry at TapPrint Sdn Bhd.

StackSpring Boot, Vue.js, PostgreSQL, Liquibase
ConstraintComplex industry-specific pricing matrices requiring seamless integration with existing tools.
ResultReduced quote generation time by 50% and significantly improved operational efficiency.
Architecture Diagram[Fig. 2]
Vue.js Client
→
Spring Boot API
↓
PostgreSQL + Liquibase

Sys.003

DOSM Data Migration System

Large-scale data migration for the Department of Statistics Malaysia at Fotia Solutions.

StackPython, PostgreSQL, Jenkins
ConstraintMigrating 10 million+ census data records with a strict zero data loss requirement.
ResultSuccessfully migrated all records, reducing manual data entry errors by 90%.
Architecture Diagram[Fig. 3]
Legacy Data
→
Python ETL Pipeline
↓
Target PostgreSQL

Sys.004

Multi-Tenant SaaS Delivery

Led a cross-functional team of 10 to develop multiple SaaS applications including eCommerce and telemedicine at Double Edge Technologies.

StackGrails, Kotlin, Vue.js, MySQL
ConstraintDelivering 5 major platforms concurrently within strict deadline and budget constraints.
ResultDelivered all 5 applications successfully, increasing client revenue by 25% in the first year.
Architecture Diagram[Fig. 4]
Cross-Functional Team
→
Standardized CI/CD
↓
5x SaaS Platforms

02. Systems

Thinking & Architecture

Design Principles

  • 01Systems must be observable. If you cannot see its state, you cannot trust its behavior.
  • 02Complexity is a liability. Isolate it, document it, or eliminate it.
  • 03Data structures define the architecture. Algorithms are just the plumbing.
*

"AI is a structural material, not just a tool. Vibe engineering matters when architecting solutions."

— Margin note, Pragmatic Innovations, 2024

Event-Driven Microservices

SKETCH.01
PUB
TOPIC
SUB

Decoupling state mutation from side effects using an append-only log.

State Machine Routing

SKETCH.02
IDLE
→
FETCHING
↓
RESOLVED

Deterministic UI transitions governed by explicit finite state machines.

03. Notes

Intellectual Fragments
N.0142024.08.12

The Evolution of Tooling

Over 14 years, the stack shifts from Grails and Android Java to Spring Boot, React, and AI-assisted coding (Codex, Cline). The underlying architectural principles—cohesion, decoupling, and resilience—remain identical.

Tags: Architecture, AI Coding, Career
N.0152023.11.05

Data Integrity at Scale

When migrating 10 million census records, zero data loss isn't a goal; it's the baseline constraint. Algorithms must be designed defensively, assuming data corruption exists at the source.

Tags: Data Migration, Python, PostgreSQL
N.0162021.04.18

Concurrent SaaS Delivery

Managing a cross-functional team of 10 to deliver 5 major platforms requires strict standardization. Shared libraries, consistent CI/CD pipelines, and unified deployment strategies are load-bearing structures.

Tags: Leadership, DevOps, SaaS