Financial Software Application Engineer at Workday

Application Developer at Workday • Current

Lead Product Manager for Jornalero Mobile App

Intern at Rebel Idealist • December 2015 - May 2016

Featured in the New York Times.

CollegeTrack Email Client

Product Manager, Developer • January 2016 - May 2016

CollegeTrack is an nonprofit organization founded by Laurene Jobs that provides resources for high school students in underserved communities to help them get into college and graduate successfully.
I was the team leader and a SaaS developer for the CTMail project, which is an email interface that allows CollegeTrack staff to easily communicate with particular subsets of their members through automatically querying their Salesforce database.

Wireless Neural Implant Research at the Swarm Lab at UC Berkeley

Research under Professor Michel Maharbiz • May 2015 - August 2015

Our goal is to create a Brain Machine Interface that is high density, long-term, and clinically viable for the human brain. Currently, neural recording is done through direct electrical measurement through conducting electrodes and potential changes near relevant neurons during depolarization events called action potentials. Problems with current technology include 1) unreliable electrical connection between active area inside brain and electronic circuits near periphery 2) practical upper bound on number of viable recording sites 3) growth of biological response degrading recording performance over time. With implanting neural dust particles on the surface of the brain, many of these problems may be solved.

Scalable Data Analytics and Machine Learning

Distributed Data Systems with Spark SQL •

Implemented a K-means ++ clustering machine learning algorithm in Spark and used it to analyze this year's political campaign data from the Federal Election Commission.

Python Concurrency Control Manager for Database

•

Created Data Structures for Strict 2 Phase Locking & Deadlock Detection. Developed Transaction Handler and Coordinator.

Q Learning Agent for Reinforcement Learning

•

Q Learning to learn values of Q States is a solid way to acquire the optimal policy for an agent when the agent is initially clueless about the world it's in (no Transition or Reward models).

qLearningDemo from Clark Chen on Vimeo.