Adrian Lam Email & Phone Number

Computer Vision Team: developing document intelligence solutions for banking applications

• Devised and optimized custom metrics to understand users' driving behavior and assess driver risk
• Performed feature engineering to transform raw telematics data into predictive variables for detecting collision events
• Coauthored white paper documenting the calculation of improved DriverScore

Boston, Massachusetts

• Developed a deep learning model to classify the mode of transportation of user trips from telematics (time series) data
• Outperformed the in-production model on test set, in both precision and recall for each class
• Decreased annual spending as a result of eliminating model dependencies on external data sources and APIs

• Created statistical models from large-scale galaxy survey data, to map out the distribution of dark matter in galaxy clusters
• Implemented object detection, shape measurement, and Bayesian inference in Python to extract lensing signal from noisy astronomical data, processing images of 500+ megapixels

Mentored and taught students through hosting exam review sessions, office hours, and substitute lectures. Developed official course materials.Phys 2030: Theoretical Classical Physics IPhys 1280: Introduction to CosmologyPhys 0790: Physics of MatterPhys 0220: Astronomy

Master Of Science (Ms), Data Science

GPA: 4.0

Master Of Science (Ms), Physics

Thesis title: Modeling a Dark Matter Halo as a Perfect Fluid in de Sitter Spacetime Thesis advisor: David Lowe

Bachelor Of Science (Bs), Physics

University Admissions Scholarship (merit-based half scholarship)