Work Experience
April 2021 - Present
Front-end Developer
Synergies
Developed Synergies Intelligence Platform - Jarvix
- Use ECharts to process various data visualizations.
- Experienced the process from Vue2 to Vue3, familiar with the development pattern of composition API and TypeScript.
- Developed a set of component libraries based on ant-design-vue encapsulation.
- Use Pinia as our state management library with flat architecture and better code splitting.
- Use design patterns and SOLID principles to improve program reusability and extensibility.
- Utilize 3D charts to facilitate Taipei MRT in monitoring passenger flow variations across different time periods and stations.
- Developed user guidelines to make users more familiar with how to use the Jarvix step by step.
- Developed autoML to improve factory productivity.
- Developed product trial version and official website registration page.
- Participate in the development of major version updates of the product and successfully release at the press conference.
Side Project
Frontend Config
Front-end configuration file using the pnpm monorepo architecture. Just add the path to the project package.json and install to enjoy common settings such as eslint, stylelint...etc. The advantage is that it is no longer necessary to write the config setting again and it can be split and isolated according to different versions of configuration files.
link here
ESlint Config
My eslint config presets extends @antfu/eslint-config
Component Library
After re-encapsulating the architecture of ant-design-vue and using vitepress to generate a static website to watch the components demo styles and test usage. Just add the path to the project.json, and install it to use encapsulated components. Maximize the reusability of common components without storing them in business project folders.
link here
Utils
Collection of common JavaScript / TypeScript utils.
- Tree-shakable ESM
- Fully typed - with TSDocs
Awards
Synergies hackathon — first place winner
Generative Feature Engineering
In the past, feature engineering often required multiple data crunches to gain insight ( analyze => gain insight => insufficient data => reanalyze... ). However, by combining GPT with our prompt tuning, we can analyze the data and produce data that meets user expectations and download and use it.
User flow:
Choose data frame
Check the column to be observed
Input data description ( optional )
Generate Feature Engineering
regenerate / save it as a new dataframe and use it.
Skills
Language
- HTML
- CSS
- SASS/SCSS
- JavaScript
- TypeScript
Libraries
- Element UI
- Ant Design
- Tailwindcss
- ECharts
- Vuex / Pinia
- Vue router
- VitePress
Frameworks
- Vue.js
- React.js
- microApp
Tools
- Vite
- Webpack
- rollup
- ESLint / StyleLint
- Git
Education
Chang-Gung University, Computer Science
Sep 2016 - Jun 2020
