Kien Duc Vu
[email protected] | +84 792121088 | linkedin.com/in/kien-vu-3318b01a7
OBJECTIVE
Data scientist with 3+ years of experience in research and development of data-intensive applications. Proficient in
modeling and processing of various data forms, including but not limited to tabular data, digital images, and natural
languages.
PROFESSIONAL EXPERIENCES
Confidential Fintech Company., Remote, Singapore June 2022 – Current
Data Scientist
• Developed end-to-end credit risk models to support underwriting, up-sale, and collection business processes.
• Developed and deployed AI solutions for various NLP tasks such as spam detection, email intent classifications.
• Built dashboard with Periscope and Sisense to monitor model performance and data integrity.
Assurant, Inc., Atlanta, Georgia, USA January 2022 – June 2022
Machine Learning Engineer
• Annotated data with open-source packages such as Roboflow
• Built and deployed in-house OCR models for license plate and odometer images reading to support car insurance
registration process.
IC3 and Marathon Oil Company Consortium, Norman, Oklahoma, USA August 2018 – August 2019
Research Assistant
Project: Formation Evaluation of Meramec Oil Field
• Classified various rock types with clustering algorithms (K-Means, GMM, Fuzzy) to identify drilling horizons
• Interpreted and predicted fluid saturation from Nuclear Magnetic Resonance with elastic regression
• Segmented and extracted minerals and organic compounds from SEM images with U-Net
Apache Oil and Gas Corporation, Norman, Oklahoma, USA May 2017 – August 2017
Geoscience Intern
• Modeling stress-dependent fracture permeability to predict horizontal well performance over its lifetime
• Applied ML clustering models (DBSCAN and K-Means) to detect possible frac-hit zones of horizontal wells
EDUCATION
Georgia Institute of Technology, Atlanta, Georgia, USA August 2019 - August 2022
Master of Science in Analytics
University of Oklahoma, Norman, Oklahoma, USA August 2015 - August 2018
Bachelor of Science in Petroleum Engineering
Courseworks: Data Structure and Algorithm, Machine Learning, Natural Language Processing, Network Analysis
SELECTED PAPERS
Exploration and Visualization of Large-Scale Pornography Uploads Metadata
• Adapted general BERT model to produce embeddings for 1.2 million porn-related sentences and keywords
• Built graph structure and visualized porn video’s links with semantics and keywords
• Detected pornographic communities having same contents with Louvain and Leiden algorithms
Publication: Vu, K., Feaser, E., Yang, J., Yu, E., Blinderman, I. (2022) Exploration and Visualization of Large-Scale
Pornography Uploads Metadata
TECHNICAL SKILLS & TOOLS
Tools: SQL, Docker, Git, Pytorch, Tensorflow, Spark, Hadoop, AWS, Azure, Databricks, Periscope, Sisense
Packages: scikit-learn, huggingface, OpenCV, tesseract
Languages: fluent English, native Vietnamese
HONORS
Vietnamese Mathematic Olympiad (2011)
Vietnamese Provincial Mathematic Olympiad (2010, 2011, 2012)
Conoco Phillips Scholarship
Departmental Excellence Scholarship for International Student
Mewbourne School of Petroleum and Geological Engineering Excellence Scholarship
University of Oklahoma Excellence Scholarship for Transfer Student
