KONG JUN SIENG

Background

Commissioned by Satoyama Farm, the Chicken Coop Monitoring System (CCMS) provides real-time monitoring of chicken coop environment and flock health, supporting the farm's digital transformation. The project adopts an edge computing + IoT + web management platform architecture, with terminal devices deployed on-site at the farm and servers managed by the client.

Contributions

• Hardware Integration — Installed Linux OS on Raspberry Pi, connected temperature/humidity sensors and camera module
• Edge Development — Wrote sensor drivers, data collection scripts, MQTT/HTTP data reporting and edge computing logic in Python
• Backend API — Built RESTful endpoints with Fast API, covering device data ingestion, data query, device management, and alert notification

Outcome

Edge system successfully deployed and validated: Raspberry Pi autonomously collects environmental data and reports to the client's server, enabling remote monitoring and anomaly alerts for the chicken coop. The complete IoT data pipeline (sensor → edge → API → visualization) is fully operational.

Background

A personal portfolio site serving as a technical validation project for AI-assisted development. Followed a 'documentation-first' workflow — completed over 10 design documents in Obsidian covering brand positioning, style guide, wireframes, and technical architecture before writing a single line of code. Then, based on those specifications, developed with hybrid AI assistance from DeepSeek V4 Flash + Ollama local models, going from zero to deployment in about 2 weeks.

Contributions

• Architecture — Designed the project directory structure, component tree, data flow, and Server/Client boundaries
• Design System — Established design tokens via Tailwind v4 @theme, integrated game-style UI components
• Full-Stack Development — Next.js 16 App Router, next-intl trilingual i18n, Zod data layer, Framer Motion animation system
• Game-style Experience — Chapter header system, blue dialog box narrative, option-button interactions, transition dialogues, intro scenario
• CI/CD — GitHub Actions lint + typecheck + build + Vercel auto-deploy

Outcome

A responsive multilingual portfolio site deployed on Vercel with an interactive game-style intro experience. The project itself serves as the best case study — demonstrating how AI-assisted + documentation-first methodology can deliver a design-driven frontend project in 2 weeks.