About me
Who is David Pak?
My name is David Pak, I am a junior studying Computer Science at
the University of Washington. I am a Full Stack developer with proficiency
in languages such as Java, Python, C/C++, HTML, JavaScript, and CSS. I have
also utilized many different frameworks for web/app development such as React,
Node, Typescript, and SwiftUI. I wish to pursue work in the software engineering,
and Machine Learning fields.
I am also an accomplished musician, tennis player, and coach.
I am currently pursuing a minor in music theory at
the University of Washington and have been playing tennis for
over ten years. Additionally, I have over three years of experience
as a recreational tennis coach
Front End
Developer
Undergraduate
student
Back End
Developer
ML Engineer
My Timeline
2022 - present
Advanced Coding Instructor - Coding With Kids
● Facilitated comprehensive workshops covering advanced coding concepts, fostering a deep understanding of web development and complex data structures.
● Mentored students up to 15 students at a time in executing real-world projects, resulting in the creation of over 200 diverse applications, including task organizers, personal websites, and social networking platforms.
● Assisted students in applying fundamental graph algorithms, enhancing their problem-solving and algorithmic thinking skills.
● Demonstrated effective communication and pedagogical skills, facilitating the understanding of complex coding principles among diverse age groups.
● Achieved a 100% 5-star review rate from parents and guardians.
2023 - present
Interpretability Researcher - AI Alignment @ UW
● Designed and implemented interpretable machine learning models to enhance transparency and trustworthiness in AI systems, collaborating with multidisciplinary teams.
● Authored research papers delving into the intriguing emergent phenomena exhibited by neural networks as they scale up in model size, training data, or the number of training steps.
● Conducted reverse engineering analysis on neural networks, specifically by discerning the underlying circuits responsible for behavior implementation within a given model.
● Created custom data visualizations and interactive dashboards to facilitate model understanding for both technical and non-technical stakeholders, effectively conveying model behavior and decision-making processes.
2020 - 2023
Tennis Coach - Jupiter Tennis Center
● Conducted clinics for groups of over 20 students, spanning various age groups, imparting fundamental tennis skills and advanced strategies.
● Provided individualized private lessons and tutoring services to students of all ages in the greater Seattle area.
● Designed and implemented personalized training programs tailored to each student's skill level and goals.
● Organized and supervised competitive practice matches and tournaments to enhance students' match-play skills.
● Focused on fitness and conditioning, incorporating drills to improve endurance, agility, and overall physical fitness.
2022 - 2025
Computer Science Degree - University of Washington
● Pursuing a Bachelor's in Computer Science at the University of Washington.
● GPA: 3.8
● Relevant Coursework: Data Structures and Parallelism, Discrete Math, Machine Learning, Algorithms, Compilers, Systems Programming, Software Design and Implemtation.
My Portfolio
Dental Health Model
A linear machine learning model that utilizes Ridge Regression to determine which factors contribute most to an individual's dental health. I am currently in the data collection process and will have a notebook published with the results by October 31st, 2023.
I Don't Know, You Choose
An interactive web application that generates a random restaurant within 15 miles of the user's location. The frontend of the application was implemented using React.js and the backend was implemented using the Flask framework written in Python. The restaurants are obtained using Google Places API's "nearby_search" endpoint, and the user's location is obtained using Geolcation.
NBA Player Comparison Tool
A program that allows users to input two NBA players and outputs a "similarity score" based on a variety of metrics. The program was implemented in Python and all of the necessary data was collected using nba_api.
NBA ShotChart Maker
A program that takes an NBA player and season as user input and outputs a shot chart graphic that displays shot selection of said player using hexagons. The program was implemented in Python and uses the nba_api along with the pandas and matplotlib libraries in Python.
Work Experience
Contact Me
Contact me here
Feel me to contact me with any questions/inquiries. Any form of contact is acceptable, however, I do prefer Gmail if possible.
: Seattle, Washington
: sungh1404@gmail.com
: University of Washington, Washington
: +1 (253)-282-1617
: English, Korean