Leader of ML/AI initiatives for Fraud Prevention at Amazon Web Services with 10+ years of experience in Data Science and risk management.
I'm a technical leader focused on building ML/AI systems at massive scale to solve complex business problems. Over the past decade, I've moved from individual contributor to strategic executive, building high-performing data science teams. I focus on technical excellence, talent development, and delivering measurable outcomes through ML solutions.
Data Science Manager, AWS Fraud Prevention | May 2023 - Present I lead ML/AI initiatives for enterprise fraud prevention while managing a team of data scientists and business analysts. We have deployed 200+ machine learning models that significantly increased automation while maintaining high accuracy. I direct AWS account compromise detection strategy for generative AI services and balance technical leadership with people management, implementing mentorship programs and cross-functional frameworks that reduced deployment time from weeks to days.
Data Scientist, AWS Fraud Prevention | May 2021 - April 2023 I contributed to building the real-time AWS account compromise prevention platform, designing ML models using deep neural networks and transformers to detect fraud patterns across AWS services. I developed a real-time classifier processing 1.1 million events per second with sub-100ms latency. My work contributed to $15M in annual cost savings for AWS through high-accuracy fraud detection with significantly reduced false positives. I established MLOps best practices and CI/CD pipelines that reduced deployment time from weeks to days.
Senior Manager, Data Science | January 2018 - April 2021 I led risk management transformation for a $32 billion investment portfolio while building and managing a team of 5 data scientists. My team developed a comprehensive risk management platform with significantly reduced false positives while ensuring regulatory compliance. I pioneered alternative data sources and advanced modeling techniques that improved risk-adjusted returns. I presented analytical findings to C-level executives and regulatory bodies, then implemented a model governance framework that was adopted across other business units.
Manager, Analytics & Insights | January 2016 - December 2017 I managed a team of 5 analysts generating insights from Fidelity's investment and risk data. I led development of risk assessment models and portfolio analytics that improved risk-adjusted returns and enhanced investment decision-making processes. My team built predictive analytics for risk management and portfolio optimization. I established automated reporting frameworks using Tableau and Python, significantly reducing manual processes. I initiated Fidelity's first machine learning pilot projects in risk management.
Master of Business Administration | May 2019 Specializations: Business Analytics, Quantitative Finance, Financial Technology
Bachelor of Science in Business Administration | December 2013 Specializations: Information Systems, Marketing
Deep Neural Networks, Transformers, LLMs (GPT, BERT), Anomaly Detection, Time Series Analysis
MLflow, Kubeflow, SageMaker, CI/CD for ML, A/B Testing, Model Monitoring, Docker, Kubernetes
Python, R, SQL, PySpark, AWS (EC2, S3, Lambda, EMR), Spark, Hadoop, Kafka, Airflow
Team Development, Technical Mentorship, Performance Management, Strategic Planning, Resource Allocation
Led development of real-time fraud prevention platform for enterprise clients, delivering $15M in annual savings through advanced ML techniques and high-accuracy fraud detection.
Lead AWS account compromise detection strategy for Generative AI services, directing cross-functional team to deliver significant fraud reduction while maintaining low false positive rates.
Built risk management platform for $32B portfolio at Fidelity with significantly reduced false positives while achieving regulatory compliance through scalable ML infrastructure.
Invented and patented systems and methods for detecting anomalous data traffic over proxy servers using isolation forest algorithms and risk-based feature extraction. US Patent 10,826,927.