David Wu Email & Phone Number

Setauket- East Setauket, NY, US

• Developed and optimized a deep learning model in Python using the PyTorch library that classifies images in Zebrafish embryos to analyze the effect of certain drugs on blood clotting disorders, achieving an accuracy of.
• Implemented data augmentation techniques and feature selection by resizing the images and calculating the Corrected Total Cell Fluorescence to reduce the sample size, increasing the Convolutional Neural Network (CNN).

Ann Arbor, Michigan, United States

• Built a recruitment application using React.js, HTML/CSS utilizing design team wireframes, constructing 10+ dynamic elements to streamline multi-round recruitment process, reducing candidate selection time by 30%
• Collaborated with a team of 6 developers to engineer frontend and backend functionalities with Google Firebase, Firestore, and Authentication services, creating an application scaling to 500+ applicants per year
• Implemented member search functionalities using Javascript, resulting in 147% faster applicant lookup times

• Integrated 50 e-commerce, accounting, and CRM service APIs such as Hubspot and Shopify with the Endgrate unified API using a Python Flask and MongoDB backend, facilitating over 10,000 client data transfers
• Developed rigorous protocol tests for transferring data to and from external APIs, achieving a 100% coverage of all default and custom API data fields for high integration reliability, ultimately exposing 30+ bugs
• Enhanced the API endpoints by modifying the returned data to include success flags, transferred data, and associated data when pushing and pulling data, improving user experience and operational efficiency by 140%

• Developed a supervised Neural Network and Support Vector Machine (SVM) model in Python to predict remission (being free of depression) in depressive patients undergoing antidepressant treatment, with a 79% accuracy
• Designed a general meta machine learning model pipeline for constructing a binary classifier using small datasets containing large amounts of features, ultimately increasing the final predictive accuracy by 10%
• Presented my methodology to over 100 doctors at the 2023 American Statistical Association Conference and co-first authored a paper published in the peer-reviewed journal PLOS One

Bachelor Of Engineering - Be, Computer Science, 3.96/4.0

Currently pursuing a Bachelor of Science with a Computer Science major and a Mathematics minor

High School Diploma, 4.2/4.0

Multivariable Calculus, Differential Equations,

Nanodegree, Data Structures And Algorithms

Took a high-level "nanodegree" course from Udacity that covered a wide data structures and algorithms. The course was project-based.

Education record

Took courses in Data Structures and Algorithms.