Nathan Ansel
Data Scientist and AI Engineer passionate about applying machine learning and advanced analytics to solve business challenges, optimize decision-making, and uncover meaningful patterns in data.
Proficient in Python and R, with a strong foundation in transforming complex data into actionable insights and building scalable, data-driven applications.
Continuously exploring emerging AI technologies and best practices to develop efficient, reliable, and impactful solutions.
Projects
CRUMS: Rethink Your Food Waste
Sep 25 - Apr 26
Led a 7-person team building a smart food-waste bin that uses Black Soldier Fly larvae to upcycle 20kg of waste a day, automated by Computer Vision and Embedded Systems.
Hospital Robot Fleet Sizing: An Agent-Based Simulation
Jan 26 - Apr 26
Engineered an in-browser hospital simulation, applying CRN across 3,500+ paired replications to isolate variance and quantify robot fleet ROI.
Predicting Consumer Preferences: Discrete Choice Models & ML Ensembles
Aug 24 - Aug 24
Built discrete choice models (MNL, Random Forest, XGBoost) in R, combined via a soft-voting ensemble, to predict consumer trade-offs among vehicle safety-feature bundles — placing 1st runner-up in a Kaggle competition.
Heuristic Scheduling for NP-Hard Assembly Lines
May 24 - Jul 24
Four scheduling heuristics benchmarked across 1,944 generated BOMs, wrapped in a React + FastAPI web app that puts near-optimal production schedules in front of non-technical production managers.
City-Scale Mobility: Modeling 2 Million Commuters using OpenCV and KDE
Sep 23 - Jun 24
Modeled commuter mobility for 2 million Phnom Penh residents using kernel density estimation, backed by an automated Selenium/OpenCV traffic-vision pipeline processing 7,200 images/hour.
End-to-End Supply Chain Optimization: An MILP Approach in Julia
Jan 24 - Apr 24
Built a Mixed Integer Linear Program (MILP) in Julia to optimize a multi-stage national supply chain across 16 fishing grounds and 100 cities.
Data-Driven Inventory Planning Management @ Tai Sin Electric
Jan 24 - Apr 24
Built a real-time Power BI inventory-planning dashboard for Tai Sin Electric, Singapore's largest cable and wire manufacturer, replacing gut-feel reordering with data-driven forecasts.
Food Security in SEA: An Award-Winning Machine Learning Model for Crop Yields
Sep 23 - Dec 23
Built and iteratively refined multiple linear regression models predicting crop yield across 10 staple crops in Southeast Asia from environmental, economic, and political factors — awarded Best Project.