Phu Tran, PhD

⬇️ PDF resume

Data Scientist and AI Engineer with extensive experience applying data science and ML/AI across diverse domains. Skilled in problem solving, data-driven problem formulation, designing and implementing ETL pipelines, and developing a range of ML algorithms, including supervised, unsupervised, and reinforcement learning. Familiar with full-stack application development and cloud deployment with CI/CD pipelines. Experience working with interdisciplinary teams, with a strong ability to communicate complex technical concepts clearly.

Skills

• Machine learning: Generative AI, Transformers & LLMs, Computer Vision, Reinforcement Learning, Supervised & Unsupervised Learning
• Programming Frameworks: FastAPI | Streamlit | Docker | NextJS | Nodejs | BigQuery | Google Cloud Platform | LangChain & LangGraph | Linux | Python | SQL | Scikit-learn | PyTorch | Ray | Weights and Biases | MLFlow | Git | DVC | CI/CD

Certificates

• Google Data Analytics Professional Certificate (completed 2021)

Experiences

Research Associate (AI Engineer), Mayo Clinic

Scottsdale, AZ | June 2025 – present

• Research, build, and deploy end-to-end agentic AI systems that retrieve patient EHR and respond to questions about a patient in different contexts.
• Develop AI-assisted solution for clinical trial matching and automation of cancer registry patient enrollment.
• Build scalable backend and frontend with UX/UI
• Tech stack: Azure OpenAI and Google Gemini APIs, Python FastAPI, NextJS, NodeJS, Docker, Google Cloud Platform (Vertex AI, Cloud Run, Cloud Build, BigQuery, Cloud SQL), LangChain, LangGraph

Postdoctoral Associate, Brandeis University

Waltham, MA | Feb 2022 – Apr 2025

• Lead multiple research projects to develop ML/AI models to forecast and control bio-inspired materials.
• Developed deep learning model (CNN, RNN) to measure velocities of object in experimental videos, significantly outperform existing rule-based method.
• Developed deep learning model (Quantized CNN Autoencoder, Transformers) to predict long-range dynamics of active materials, achieved prediction of 2 times longer into the future comparing to existing method.
• Developed reinforcement learning framework (PPO algorithms, CNN-based policy network) to control active materials in simulation and experiments.

Research Fellow, Nanyang Technological University

Singapore | May 2018 – Jan 2022

• Led projects to develop ML algorithms for predicting aircraft locations using satellite data and AI agent to assist air traffic controllers. Transformed operational requirements to high impact research questions, with 02 large datasets collected.
• Achieved 30% improvement in aircraft trajectory prediction by data augmentation and infusion, combined with enhanced GRU network architecture.
• Developed Human-AI (reinforcement learning for assisting human air traffic control) user interface software, resulting in 01 software prototype and 01 public demonstration in the prestigious Singapore Airshow 2020.
• Initiated and deployed a large MySQL database server in Linux (1.6 billion rows of time series data), serving 20 researchers.

Services and Other Achivements

• Publication Chair of The 1st International Conference on Artificial Intelligence and Data Analytics for Air Transportation (AIDA-AT 2020)
• Published 20+ peer-reviewed research articles, with 300+ citations on Google Research Scholar

Education

• Nanyang Technological University, Singapore, PhD in Mechanical Engineering | Aug 2012 – July 2017

• University of Technology, HCMC, Vietnam, BEng in Mechanical Engineering | Aug 2007 – Apr 2012