Amy Yu
  • Projects
  • Experience
  • Education
  • Résumé

Amy Yu

Purdue University Computer Graphics | Meta Software Engineer | University of Waterloo Computer Science

Projects

Keyframes AI

2025 | Purdue University

A web application that uses diffusion models to generate stop-motion animations from figurine photos and a series of poses. Approved as my Master's Thesis.

Key Concepts: Generative AI Model Finetuning, Pose Detection, UI/UX, Creative AI

Read Thesis Slides Demo 1 Demo 2

A11yShape

2024 - 2025 | Purdue University

An accessible web application that uses GPT-4o to generate code and descriptions of OpenSCAD models to help blind users understand, generate, and verify 3D models. I worked on the UI design, Backend Implementation, OpenAI API Integration, Prompt Engineering, and User Study Design. Paper published in ASSETS '25.

Key Concepts: LLM Image Captioning, Prompt Engineering, OpenAI APIs, HCI, Accessibility

Read Paper Preview Video Demo Video

Manga OCR

2024 | CS593 Computer Vision Project

A web application that automatically translates manga pages from Japanese to English by detecting speech bubbles and extracting text.

Key Concepts: Computer Vision, OpenCV, OCR, Model Evaluation

Details Source Code Project Report

Sea Creature Classifier

2024 | Purdue University

A ResNet-50 model trained on images of sea creatures so it could label a user inputted image with the creature's scientific name. I worked on creating the training dataset using a webcrawler to extract representative images for each species.

Key Concepts: Computer Vision, ML Model Training, Data Extraction

Try It

FathomGPT

2023 - 2024 | Purdue University

A system for understanding ocean science data via a natural language interface, using OpenAI's LLMs to enable the creation of complex queries to analyze data. I worked on the RAG system for name resolution and semantic search, the scripts for filtering high-quality images from the sea creature dataset for model training, and the pre- and post-processing stages for the final output. Paper published in UIST '24.

Key Concepts: RAG Systems, LLMs, OpenAI APIs, Agentic AI, Data Pre- & Post-Processing, Data Engineering

Read Paper Preview Slides Demo Video

Batch Mentions

2022 - 2023 | Facebook

Added a Facebook Group mentions feature that allows users to mention multiple people at once and send notifications to all members of a Facebook Group (millions of users). Worked with UX designers and other teams (eg. Notifications Team) to launch the feature.

Key Concepts: Scalable Systems, Cross-Functional Collaboration

Structured Typeahead Suggestions

2020 - 2022 | Facebook

Determining high confidence typeahead suggestions and decorating them with images and snippet. These suggestions are selected based on entity features, such as degree of connection, recent interactions, and popularity.

Key Concepts: Ranking Models, A/B Testing, Large-Scale Data Pipelining

Instrumentalize

Winter 2019 | CS489 Computational Audio Project

A command-line tool to transcribe a piece of music from one instrument to another.

Key Concepts: Computer Audio, Pattern Detection, Data Processing

Project Report

Forecaster

Fall 2018 | Facebook Internship

Internal tool for timeseries extrapolation using Prophet.

Key Concepts: Machine Learning Models

R2RDump

Spring 2018 | Microsoft Internship

An open-source command-line tool that parses .NET ReadyToRun binary executables to display assembly code, ARM and x64 UnwindInfo, and garbage collection tables.

Key Concepts: Low-Level Programming, Assembly Code, Compilers

Details Source Code

Animator

Winter 2018 | CS488 Computer Graphics Project

An OpenGL application that allows users to create animations by specifying keyframes. The models will behave according to some laws of physics, such as gravity and the conservation of momentum.

Key Concepts: Computer Graphics, OpenGL

Details Project Report

Project Panama

Winter 2017 | IBM Canada Internship

Native method handles for calling C/C++ methods from Java programs. Added the new feature for the IBM's J9 virtual machine which is implemented differently from Oracle's version of OpenJDK Project Panama.

Key Concepts: Low-Level Programming, Function Call Stacks

Details

Kapsel Fingerprint Unlock

Spring 2016 | SAP Canada Internship

Integrated the Android 6 Fingerprint API into the Kapsel Logon plugin so that users can authenticate with their fingerprint instead of a passcode. This feature was launched to production. More info.

Key Concepts: Hybrid Mobile App Frameworks, API/Tool Integration

Octane Template Binding

Spring 2016 | SAP Canada Internship

Prototype of a tool for generating HandlebarJS templates used when creating index cards with SAP Octane. Provides an interface for replacing the handlebar names with the field names of the user's JSON data. Presented at a meeting and launched to production.

Key Concepts: String Manipulation, Data Extraction

Millionaire Labs Prototype

2015 | Early Stage Startup

An educational bank simulator web application with PHP for the back-end, Bootstrap for the front-end, and MySQL for the database. I quickly created the MVP from start to finish based on ideas and feedback from the team.

Key Concepts: Full-Stack Development, UI Design, Prototyping

Details

Transpost

2013 | Freelance Project

I was hired by a real estate agent from Red Carpet Real World Realty Inc to develop a Java application that extracts real estate information from webpages and uses Selenium webdriver to automatically fill the HTML form of an advertising website.

Key Concepts: Webdrivers, Data Extraction

Details

ECOO Contest Solutions

2013 | Grade 12 Computer Science Class Final Project

A website that sorts, categorizes and displays ECOO (Educational Computing Organization of Ontario) contests from the last 30 years. I mainly worked on the PHP back-end, and was in charge of the part that programmatically translates Turing solutions into Java.

Key Concepts: Backend Development, Compilers

Details

Experience

Research Assistant

Purdue University

August 2023 - December 2024
Worked with Professor Angus Forbes and Professor Liang He

Software Engineer

Meta (Facebook)

July 2019 – August 2023
Search Typeahead Team

Software Engineering Intern

Facebook

Sept 2018 – December 2018
Data.AI Team

Software Engineering Intern

Microsoft

May 2018 – August 2018
.NET Runtimes Team

Software Engineering Intern

Cisco Meraki

September 2017 - December 2017
MR Features Team

J9 Virtual Machine Developer Intern

IBM Canada

January 2017 - April 2017

Core Infrastructure Developer Intern

SAP Canada

May 2016 - August 2016

Web Developer Intern

Bank of Canada

September 2015 - December 2015

Education

Purdue University

Master of Science (Computer Graphics)

August 2023 – December 2025

University of Waterloo

Bachelor of Computer Science (With Distinction)

September 2014 – June 2019

Machine Learning

Python • Generative AI • LLMs • Computer Vision • NLP • ML Training Pipeline

Low-Level Developer

C++ • Assembly • Embedded Systems • Real-Time Programming

Back-End Developer

PHP • Ruby on Rails • SQL • Ranking/Recommendation

Oracle Certified Java Programmer

OOP • Swing Library • Hybrid mobile app development
Copyright © Amy Yu 2015 - 2025
  • Download Résumé