Stanford University

Massachusetts Institute of Technology

Data Scientist at Lacework

Worked in a team that develops and maintains data-processing components in a cloud security product.  Added functionality to adjust properties of alerts in the product for anomalous behavior, which affect the prominence with which the alerts are displayed in the UI, based on historical data.  Implemented optimizations in an anomaly detection system to increase the maximum amount of data that can be processed by the system.

Lacework

Software Engineer

Continued to work on an analytics system for an Internet of Things security product.  Implemented a module for combining related security events that occur at different times to add context to generated alerts.

(Palo Alto Networks acquired Zingbox in September 2019.)

Palo Alto Networks

Machine Learning Engineer

Worked on an analytics system for an Internet of Things security product.  Developed infrastructure and logic for alert generation based on data about the network traffic of devices.  Implemented aggregation and feature extraction functionality to collect features used in models of network traffic.  Worked on anomaly detection to identify unusual behavior in device communication.

Zingbox Inc.

Data Scientist/Engineer

Worked in a team focused on display advertising for mobile devices.  Maintained and developed a system for classifying publisher web sites as mobile-friendly or not.  Worked on projects involving the targeting of ads for mobile requests, using comparisons of mobile and desktop requests to identify differences between the types of requests.

Google

As part of a system that automatically builds machine-learning models for detecting abuse in user-generated content, developed a module for extracting features from content samples.  Created grid scripts for collecting content submission histories for users that are inputs in the calculation of content creation reputations.  Evaluated the effectiveness of different machine-learning models for abuse detection trained using various data sets.  Implemented a module that classifies user-generated content on the basis of user reputation information obtained from abuse reports.

Yahoo!

Developed and implemented desktop sharing functionality for a video collaboration system with Java and Flash components.  Worked on a method of delivery of videos to users.

ViVu Inc

Explored approaches for developing recommendation systems.  Implemented, in Visual J#, and evaluated several techniques for providing movie recommendations to users based on past movie ratings.

Microsoft Research

Research Intern

Designed and implemented, in Java, a set of system components for writing metadata to database storage in a service-based distributed system.

IBM Almaden Research Center

Designed and implemented services to add functionality to a Java-based distributed video management system that automatically ingests video and allows that video to be retrieved.

Damon Mosk-Aoyama