Hi :)
I'm Gina.
I build data-driven products and predictive models that turn raw information into real decisions.
About
Passionate about predictive analytics, model explainability, and production-ready data solutions. I like working where data engineering, model development, and stakeholder dashboards meet.
Skills
Programming & Data
Collection, cleaning, pipelines, processing
Analysis & Visualization
Trends, hypothesis testing, dashboards
Machine Learning & AI
Predictive models, pattern recognition
Software Engineering
Reliable, scalable applications
Experience
Data Science Intern
VPBank · Digital Bank Team- Analyzed 5M+ customer profiles to build a Credit Card Approval prediction model.
- EDA, feature engineering, and benchmarking with DeepGBM, SAINT, and IGTD.
- Built PowerBI dashboards for 10+ analysts and stakeholders.
- Collaborated on AI-driven risk scoring and fraud detection research.
Data Analyst
Greentech Vina · Data & Analytics Team- Consolidated 10M+ international trade records with Excel and Python.
- Analyzed market anomalies and customer segments in Tableau.
- Improved reporting pipeline efficiency by 30%.
AI Engineering Intern
Vietnam Academy for Science and Technology- Processed 1.5M+ air-quality records via time-series analysis, improving anomaly detection by 12%.
- Built Looker Studio dashboards tracking long-term pollution trends.
- Ran A/B tests evaluating pollutant levels pre/post intervention.
Head Math Teaching Assistant
Colby College · Waterville, ME- Led study sessions for 30+ students in Linear Algebra and Statistical Inference.
- Coordinated grading and academic feedback with faculty.
Projects (6)
ML & Statistics
FutureProof: Bankruptcy Detection
Predictive models on ~7,000 companies using financial metrics with SHAP-based interpretation and a live Streamlit dashboard.
ML & Statistics
CityFlow AI: Real-Time Tracking
YOLOv8 pipeline at 79% mAP@50 with Deep SORT + BoT-SORT multi-object tracking and a traffic analytics layer.
ML & Statistics
AdoptGraph: Social Tech Forecasting
Technology adoption forecasting across 100K+ Facebook users via a 3-layer Graph Convolutional Network (~80% accuracy).
Full-Stack
ColbyNow Merchandise
Campus marketplace with 17+ production features, semantic search via SentenceTransformer, CI/CD, and ~95% test coverage.
Full-Stack
Mulevents: Campus Event Platform
Full-stack Flask app for Residential Life with QR-code attendance, Google Calendar integration, and Heroku deployment.
Data Engineering
PharmaGuard
MySQL schema unifying drug details, recalls, and adverse events. Python ETL pipeline with Pandas, Spark, and Hive.
Contact
Open to projects, new-grad roles, and collaborations.