New York City
Building server infrastructure for next-generation web apps.
Responsibilities include testing quant strategy hypotheses and writing production model code. Built data pipelines and front-end tools for tracking the impact of ongoing events on markets. Developed curriculum and teach quarterly sessions of an engineering course about D3.js. Python, Java, Spark
Worked under Search, on a team responsible for computing weights on the knowledge graph. Contributed to maintenance of offline data pipelines for graph data, as well as implementing latency-sensitive ranking code that runs within the search request/response cycle. C++, Java
Developed tools for detecting botnets and ad fraud hidden in real-time ad market data. Designed and implemented a model for uncovering dense clusters in terabytes of ad market data, based on the MinHash probabilistic data structure. Python, Scikit Learn
Consulted on data projects under the name Bit Aesthetics, including for a major US TV network. Developed a system for detecting trending n-grams in streaming text data based on the Apriori algorithm. Python, MongoDB
Contributed to internal data infrastructure projects as well as open-source projects in the Hadoop family. Built several self-initiated visualization projects, including “Visualizing Facebook Friends”, which appears in the IPO filing. Java, R, PHP
Server and client library for sending Web Push (VAPID) notifications, with Keras and Jupyter integration. I host a public server instance which has relayed over five million messages on a shoestring budget.
Chrome and Firefox extension for visualizing the tree structure of Twitter conversations, with thousands of active installs.
Harvard Business Review, July-August 2011 issue (pages 32 and 33). Visualization of Groupon revenue breakdown.
Created a visualization of a sample of ten million friend pairs to construct a world map. Published by Facebook and featured by the BBC and The Economist.