Po-Han Chen (陳柏翰)

Experience

1/2022 - Present

Migrated a legacy React project to Next.js, incorporating Redux, Redux Saga, and modern CSS.
Developed critical pages including the user dashboard, company page and payment pages.
Implemented new features such as contact import from other services and cross-device auto-logout mechanism.
Enhanced website performance through server-side rendering, caching services, and continuous optimization using tools like PageSpeed Insights.
Resolved challenging errors, such as an issue with Algolia during the server-side rendering phase that led to a Next.js server crash.
Enhanced team's data analysis efforts by integrating analytics tools and event tracking systems.
Spearheaded the integration of Strapi (a content management system), enhancing the team's ability to quickly produce pages and distribute content.
Pioneered the integration of LLM model API services.

11/2019 - 4/2020

Conducted research in deep learning-based Natural Language Processing (NLP), exploring innovative approaches in machine learning.
Experimented with various training methods and models, including multi-task learning and language models such as BERT.
Authored a publication titled "Detecting Deceptive Language in Crime Interrogation".
Investigated Memory Networks and applied them to develop an application for detecting deceptive language in text-based games. This involved a collaborative effort with an Israeli research team from a university in the United States.

1/2018 - 6/2018

Spearheaded the research and development of Ethereum smart contracts, focusing on contract coding, Dapp infrastructure, and security analysis.
Implemented the frontend user interface using JavaScript, HTML, CSS, and Bootstrap.
Integrated backend APIs with the frontend.

6/2016 - 1/2017

Contributed to the PowerPoint Online project, focusing on the feasibility study of replicating desktop app features in the web version, particularly the upload of GIF images.
Wrote unit tests using Selenium, contributing to the robustness and reliability of the codebase.
Utilized JavaScript and CSS to develop and enhance web application features, with a working knowledge of C#.
Independently developed and implemented an interactive web-based game using HTML Canvas, incorporating Microsoft Cognitive Services for advanced facial and emotion detection.

2013 - 2018

React and Next.js
TypeScript and JavaScript
CSS
Cloud services, including AWS and Heroku
Content Management System development (specifically Strapi)
Web Event Tracking and SEO Techniques
Skilled in integrating and utilizing services and APIs such as Algolia, Google Analytics, Mixpanel, and Sentry.