Hey there 👋

I'm Siyuan Peng

I am a software engineer, a full stack developer, and a Computer Science and Maths double major.

Know MoreContact Me
Powered by GatsbyJS, ReactJS, TailwindCSS & Vercel

.skills()

From web development to machine learning, from frontend to backend, from JavaScript to Python, I use whatever tools and technologies I need to make the world a better place.

Frontend Development

Where creativity is visualized.

  • HTML 5
  • CSS
  • JavaScript
  • TypeScript
  • ReactJS
  • GatsbyJS
  • TailwindCSS
Backend Development

Where services are built.

  • NodeJS
  • MongoDB
  • Django
  • Java Servlet
  • Python
  • REST API
Cloud Service

On the edge of the cloud.

  • Heroku
  • Google Cloud Platform
  • DataStore & App Engine
  • Twillo
Data Science

Where insights are gained.

  • OpenCV
  • Machine Learning
  • NumPy
  • Biometrics

const projects = () => {

PingMe Google Software Product Sprint

A web application that builds a bridge between minority groups and those who love to help. Users can report emergency situations or request non-emergency assistance. Users nearby can then see the reports on map and help out those in need. This is helpful for people who are physicaly disabled or under threat to dictate emergency situation over 911.

Features

  • Report emergency situations with location by one tap.
  • Request non-emergency assistance with location by a HTML form.
  • Visualize nearby requests on Map with location and messages.
  • Options to offer assistance or contact the requester for more details.
PingMe
Full StackTailwindCSSJavaScript
NodeJSJava ServletGoogle DataStore
Google Cloud PlatformGoogle AppEngine

ColbyEat
Personal MERN Full Stack Project

ColbyEat is a full stack MERN web application that gives a better experience of checking dining menus. It scrapes the raw and ordinary dining hall menu and notifies users of daily menu in a beautiful text message. It allows users to store dietary and location preferences so that they can receive notifications of dish items that fit their preferences.

Features

  • User can sign up and log into the application with Email.
  • Setup profiles with meal and dining hall preferences.
  • Receive text message of daily menu based on preferences.
  • Options to adjust notification time and frequency.
A dinosaur
Full StackHeroku
TailwindCSSReactJSJavaScript
NodeJSMongoDBDjangoExpressJS

Checklist CodeDay Labs

A full stack web application that gives better experience in automated checklist creation. It allows users to create checklists from templates with customized inputs. It also allows users to share their checklists with others. All data are secured in MySQL database.

Features

  • Fast response time using RedwoodJS for full stack.
  • Create checklist from a template.
  • Manage, edit checklists in modern user interface.
  • Secure sharing of checklists.
CodeDay
Full StackVercel
TailwindCSSReactJSRedwoodJS
PrismaMySQLGraphQL

Java IDE
Object Oriented Design Course Project

A multi-platform Java Integrated Development Environment for editing and compiling Java source code. Developed following the MVC and Object Oriented Design principles by a group of 3. Apart from supporting the following useful features, it also integrates our own implementation of Java compiler and standard input/output.

Features

  • Supports find & replace, toggle comments, bracket matching
  • Integrates pretty print formatting and automatic syntax hightlighting
  • Compiler and error checking & standard input/output on Terminal
  • Includes our implementation of tokenizer, parser, semantic analyzer
Java IDE screenshot
Object Oriented Design
Desktop AppMVC Pattern
CSSJavaFXJavaCompiler

Eye Movement Biometrics
HCI Research Project

Eyes, the passwords to your devices -- by combining eye movement data during natural text reading with cognitive and linguistic features, I build supervised learning pipelines that improved the accuracy of eye-movement biometrics from 60% to 73%. Results are published in ACM Symposium on Eye Tracking Research and Application 2022 (Best Technical Abstract Award).

Outcomes

  • Utilized OpenCV, random forest and neural network to identify users
  • Trained supervised machine learning models and built pipelines
  • Improved the accuracy of biometric identification to 73% by using a combination of textual data and cognitive factors
Java IDE screenshot
Research ProjectSupervised Learning
ML PipelineOpenCVNumPy
Scikit-learnPandasMatplotlib

.experience()

With strong experience in Python and data science, I was selected in freshman year to conduct one-year CS research. With fruitful results in eye movement biometrics, I discovered my profound passion for software development. Since then, I've been learning and working on web technologies, and used my creativity and skills to built projects that help people live better with technology.

  1. Software Engineering Intern @ CodeDay Labs

    • Collaborated within a team to design and implement an automated Checklist web application in ReactJS, Prisma, GraphQL, and RedwoodJS.
    • Responsible for frontend development, conducted code review, made 5 version iterations, supervised by Principle Engineer at Microsoft
  2. Software Engineering Participant @ Google Software Product Sprint

    • Design and developed a full stack web application PingMe that reports emergency faster in Java Servlet, JavaScript, TailwindCSS, Google Cloud Platform, App Engine, and Maps API. Collaborated within a team of four, supervised by Google Engineer.
    • Contributed to open-source in Git, integrated data storage, lead distributed development, and conducted code review and testing.
    • Practiced technical interview with Google Coaches, trained data structure and algorithms, received feedback from mock interviews.
  3. Certificate in Advanced Technical Interview Prep @ CodePath

    • Reviewed and practiced advanced data structures and algorithms to prepare for technical interview, four hours lecture and many hours of praticing per week.
  4. CS Research Assistant @ Software Engineering and Human Factors Laboratory

    • Improved eye tracking biometric accuracy from 60% to 73% by building machine learning pipelines on eye movement data with cognitive and lingustical features. Best Technical Abstract Award at ETRA 2022.
    • Reduced gaze correction time by 28% by leading and developing an semi-automated correction tool with machine learning algorithms, Fix8. Submitted to ETRA 2022.
    • Improved post-experiement data processing efficiency by 40% less time by designing and developing an Eye Tracking Extension on VS Code that calculates association between eye movement data and code.
    • News: Can Eye Movement Unlock New Levels of Security?
  5. Bachelor in Computer Science & Mathematics Double Major @ Colby College

    • Courses: Data Structure & Algorithms, Data Analysis & Visualization, Object Oriented Design, Neural Network, Analysis of Algorithms, Computational Theory, Programming Languages, Discrete Math, Probability, Linear Algebra
    • Awards: Davis Scholar, Presidential Scholar, Dean List
    • Organizations: Colby Data Science Club

.contact()

Have an open position, want to get in touch or just say hello? I'd love to hear from you.

Email: siyuan.rpeng@gmail.com

Phone: +1 (207) 660-8703

Send me a message

I will get in touch in 48 hours. Thank you!