Senior Software Engineer / Data Scientist

I lead engineering work for Visa Account Attack Intelligence (VAAI), building machine learning systems that analyze more than 300 million credit-card transactions per day to identify fraud and account-enumeration risk across e-commerce payments. I built and productionized unsupervised generative models in PyTorch, including variational autoencoders, to detect anomalous transaction behavior and emerging fraud patterns, achieving a 95% true-positive rate for enumeration attacks.

I developed Bayesian risk models in Hadoop that identified high-risk transaction and account patterns and contributed to more than $10 million in fraud savings for Visa clients. I engineer and maintain end-to-end Kafka and Spark pipelines using Python, Scala, and SQL; built internal attack-investigation dashboards with Angular, Spring, Java, and MySQL; and spearheaded active-active architecture that reduced downtime by 50%. I also brief risk operations leadership and business stakeholders on fraud trends and translate model insights into actionable strategy.

I led an R&D initiative to automate field extraction from birth certificates using supervised machine learning and computer vision. I designed a custom Python pipeline using Google Cloud Vision API and OpenCV to tokenize scanned certificates and trained classification models on more than 1,000 documents, achieving over 80% accuracy across token categories.

I applied named entity recognition research to recognize semi-structured legal-document fields. I trained and evaluated MLP, RNN, Random Forest, and SVM models using TensorFlow and scikit-learn, and performed exploratory analysis and visualization with Pandas, Seaborn, and Matplotlib. The delivered infrastructure supported long-term efficiency improvements for processes that had relied on manual review.

I implemented and tested deep neural network models to support UCSF researchers in identifying instances of heart disease.

I worked with a dataset of 10,000 electrocardiogram videos as part of the undergraduate research project.

Relevant coursework included Machine Learning, Generative AI, Deep Learning, Natural Language Processing, Artificial Intelligence, Probability, and Statistics.

Cumulative GPA: 3.54. Relevant coursework included Machine Learning, Neural Networks, Database Systems, Data Science Principles, Artificial Intelligence, Probability, Statistics, Data Structures and Algorithms, Business Analytics, and Linear Algebra/Discrete Math.