Full Stack Developer | Tech Enthusiast | Building practical system
Frontend: React.js · Next.js · SvelteKit · Flutter · Vite
Backend: Django · Flask · Spring Boot · Node.js
Databases: MySQL · Firebase · Supabase
Tools & DevOps: Git · Docker · Podman · Vercel · Vitest
-
Email: kennethongkp@gmail.com
-
LinkedIn: Kenneth Ong Kuan Phing
-
Website: kennethong02.github.io/
Full-stack app that analyzes calisthenics hold videos using MediaPipe pose detection, geometric scoring, and LLM-generated coaching feedback.
- Upload a video and select the move
- Frames are extracted at 2 FPS and run through MediaPipe pose detection
- The longest stable-pose window is identified
- Geometric metrics (angles, alignment, lean) are scored against calibration curves
- An LLM generates coaching feedback from the scores
- Results are displayed with a skeleton overlay on the best frame and a score trend chart
Moves: planche · front lever · back lever · handstand
Stack: FastAPI · React · TypeScript · MediaPipe · OpenCV · Claude/OpenAI/Gemini (configurable)
A minimal, fast personal site built with SvelteKit — doubles as a knowledge base powered by RAG. My Obsidian vault is indexed and embedded, so visitors can chat with my notes and surface connections across everything I've written. Think of it as a searchable second brain, publicly accessible.
Stack: SvelteKit · TypeScript