About Me
I'm Duy Nguyen, MS Data Science student at Seattle University (starting Fall 2025), with a track record of building production ML systems that deliver measurable business impact.
I'm seeking Summer 2026 data science or AI/ML engineering internships where I can apply my production experience while continuing to develop expertise in causal reasoning and scalable analytics.
Featured Projects
MOSAIC - AI Immigration Chatbot
Built an AI-powered chatbot with SFU Blueprint for MOSAIC, serving 660,000+ users across Canadian immigration services. Recognized as Top 4 in SFU CS Diversity Award.
- Analyzed 660K+ user interactions to build structured query categorization system
- Designed knowledge graph (Neo4j) mapping programs, services, eligibility criteria
- Built validation pipeline achieving 90% accuracy with full lineage tracking
- Partnered with legal/operations teams to codify domain requirements
UC Berkeley ML/AI Capstone - Hospital Resource Optimization
Selected as Capstone Project Exemplar for UC Berkeley's ML/AI certification. Developed predictive models projecting $30.4M annual savings in hospital resource management.
- Built neural network achieving 80% accuracy predicting patient length of stay
- Analyzed 180K+ patient records to identify key predictive factors
- Quantified business impact: $30.4M savings vs baseline approach
- Integrated AI chatbot for stakeholder decision-making guidance
NASA Flight Data Analysis - Aircraft Fuel Optimization
Analyzed 1.88 million NASA flight recorder measurements to identify fuel consumption drivers. Findings challenge industry conventional wisdom about flight optimization.
- Achieved 95.9% predictive accuracy (R² = 0.959) across 312 flights
- Discovered engine performance explains 64.4% of fuel variance—2.2x more than flight planning
- Applied ANOVA, nested F-tests, variance decomposition, and interaction analysis
- Provided evidence-based recommendations prioritizing engine monitoring
Technical Skills
Languages & Tools
ML & Deep Learning
Specialties
Other Projects
Garbage Classification - Deep Learning
94% accuracy, 100% minority class recall. ResNet34 transfer learning with live demo.
Duy Integral Theorem - ML Theory
Novel mathematical framework for understanding generalization in neural networks.
SFU Faisal Lab - Medical RAG
RAG system translating natural language to JSON for CT/MRI scan retrieval.
AI Agent - ML-Business Alignment
Agent bridging ML teams and business stakeholders for strategic alignment.
Algorithm Learning Tool
Interactive visualization for mastering tree algorithms with AI feedback.
Programmatic Business Card
Print-ready business cards built with HTML/CSS/JS. Dynamic QR codes, professional design.
Contact
Email: dcnguyen060899@gmail.com
LinkedIn: https://www.linkedin.com/in/duwe-ng/