Venus Chen

Pittsburgh,PA

Summary

Full-stack developer and applied ML builder: I ship Python/Django systems end-to-end (Azure App Service, Docker, CI/CD) and build ML workflows from messy data to clear evaluation. I’ve delivered research software in education, worked on healthcare ML for type 2 diabetes screening (manuscript under review), and I’m currently beta-testing VenusLab, a consumer recommendation app preparing for launch.

Overview

2026

years of professional experience

Work History

Research Assistant

University of Pittsburgh ，LRDC

Pittsburgh, Pennsylvania

08.2025 - Current

Worked on an AI-assisted education research application grounded in neurocognitive and learning science, translating abstract research hypotheses into usable interaction flows and UI prototypes for empirical testing.
Developed and maintained the full stack (Django), supporting iterative AI/UX experiments and ensuring system stability as research questions and model assumptions evolved.
Managed deployment and experiment releases via Azure App Service and GitHub Actions CI/CD, using Docker to guarantee reproducible environments across development and data collection phases.

Authored clear technical and experimental documentation to support collaborative AI research workflows and reduce single-developer dependency.

Research Intern

Carnegie Mellon University & Shanghai Jiangtong Un

Shanghai, China

01.2025 - 08.2025

Developed and validated ML pipelines to improve screening reliability for type 2 diabetes, optimizing for recall to minimize missed-positive cases.
Implemented data preprocessing for clinical tabular data (imputation + consistency checks), benchmarked LR vs RF, and used SMOTE to mitigate class imbalance; evaluated via AUC, recall, and calibration-aware trade-offs for deployment-like screening use.
Consolidated experimental results into publication-grade figures/tables and contributed to Methods and Evaluation sections; optimized RF reached Accuracy 0.807 / Recall 0.834 / AUC 0.873.
Co-first author, manuscript submitted/under review.

VenusLab — Beauty Product Matching App

Self project

Building a consumer-facing recommendation app that personalizes beauty + fitness suggestions based on user context and preferences; currently in beta preparing for launch.
Prototyped AI-assisted food recognition (photo → meal category/estimated nutrition features) and used outputs to drive practical recommendations (workout prompts, skincare/beauty routines).
Implemented core data handling and recommendation logic; iterated rapidly through debugging and user testing feedback.
Structured the project for shipping with modular components, version control, and a release checklist; produced launch-ready assets and lightweight documentation for maintenance and expansion.

Education

Bachelor of Science - Data Science & Computer Science Minor

University of Pittsburgh

Pittsburgh, PA

05-2027

Skills

Programming: Python (pandas, NumPy), Java, SQL (basic), Git

ML/AI: data cleaning, feature engineering, model eval (Precision/Recall/F1, AUC), class imbalance (SMOTE), food image recognition (feature extraction)

Models: Logistic Regression, Random Forest

Web/Deploy: Django (full-stack), Azure App Service, Docker, GitHub Actions (CI/CD)

Product/Research: UI/UX iteration for AI/education research tools; shipped user flows from vague requirements; beta testing documentation

Timeline