Computer Vision
Health App Development
Real-time food nutrition analysis to help users track dietary intake from meal images.
AI Engineer + Backend-Heavy Full-Stack Developer
Tien Duong
I build end-to-end AI systems across RAG workflows, database design, REST APIs, data pipelines, and computer vision, turning research ideas into reliable application infrastructure.
From telescopes to artificial intelligence
Research, model building, data pipelines, and teaching.
AI
RAG
Computer Vision
Backend
APIs
Databases
DevOps
Cloud
Deployment
About
I am an AI engineer and backend-heavy full-stack developer building RAG architectures, database schemas, REST APIs, data pipelines, and production maintenance workflows.
M.S. Computer Science
California State University, Fullerton · May 2026
B.A. Astrophysics and Applied Mathematics
University of California, Berkeley · Dec 2022
AI Application Systems
RAG workflows, LLM API integration, retrieval pipelines, and model-backed features.
Backend Infrastructure
REST APIs, PostgreSQL schemas, indexing, cron jobs, and cloud operations.
Computer Vision Research
RGB-depth modeling, nutrition estimation, medical image generation, and evaluation.
Projects
Full Stack + AI
Coach Scheduler App
AI-assisted platform for lesson planning, schedule management, parent outreach, email reply handling, administrative workflows, and earnings tracking.
Publications
VRAM Optimization for Transformers Based Nutrition Prediction Model
Proposed and evaluated memory-efficient training techniques for a transformer-based nutrition prediction framework, reducing peak VRAM by 77% and reserved VRAM by 78% while maintaining model accuracy.
DOI: 10.5281/zenodo.20048116
Experience
Jan 2026 - Present
ReliBon AI, Full-Stack and AI Engineer
- Designed backend infrastructure for a real estate agent AI assistant, including relational database schemas, REST APIs, service interactions, and AI integration workflows.
- Built data ingestion pipelines to extract, clean, transform, and prepare structured data for application use and AI-powered retrieval.
- Managed cloud database operations, cron maintenance, and indexing/query optimization that reduced execution time from 35 seconds to 5 seconds.
Mar 2025 - Present
Fufu Academy, C++ Instructor
- Designed and taught structured C++ curriculum covering variables, control flow, arrays, and algorithmic problem solving.
- Created hands-on coding projects and adapted instruction for diverse learners, including students requiring accessibility accommodations.
Sep 2024 - May 2026
California State University, Fullerton, Computer Vision and AI Researcher
- Developed and optimized multimodal machine learning models integrating RGB and depth data for nutrition prediction.
- Reduced VRAM consumption of a transformer-based nutrition prediction model by about 80% while maintaining accuracy, with evaluation using AUC and F1-score.
Jan 2025 - May 2026
California State University, Fullerton, Teaching Associate
- Developed Python curriculum covering core programming concepts, data structures, OOP, debugging, and code formatting.
- Designed coding assignments with automated unit tests and structured evaluation workflows for correctness, efficiency, and code quality.
UC Berkeley
Astrophysics Research Assistant
- Worked on supernova and gamma-ray burst detection using Chandra Telescope and Young Supernova Experiment data.