During my year in the accelerated Master's program at UC Berkeley, I helped teach and facilitate two separate classes. I loved teaching so much that I was actually excited to get up at the crack of dawn every Friday and lead my early morning lessons. The first course focused on discrete event simulation where I guided students through event graph modeling and sensitivity analysis using the software package SIGMA. The second course covered mathematical modeling and output analysis using Excel Solver.

Before starting graduate school, I spent a summer working for Boeing at the final stop on the assembly line, the Everett Delivery Center. I crawled through all kinds of wide-body planes, saw the innumerable parts as they were assembled together, and redesigned workflows to enhance how those parts were tracked. I created automated, visual tools in Microsoft Access and Excel to improve identification of FAA controlled components and reduce vital reporting errors. Additionally, I worked on integrating RFID identification and tracking of parts into the production process.

The Laboratory for Automation Science and Engineering, lead by Ken Goldberg, is a research group focused on bringing automation and robotics into our daily lives. I contributed to a group investigating new systems of brachytherapy cancer treatment using optimized delivery channels in 3D printed, custom fitting molds placed around tumor areas.

I've had a long and mostly healthy obsession with the game Counter Strike: Global Offensive. Recently, I implemented a web scraper using Java and SQL to aggregate and store statistics on 18,000 professional matches from the site HLTV. This data is fed into an R based model that incorporates Elo rankings and additional match information to predict future outcomes. Further work will include additional input variables such as individual player performance and fractional rosters to improve prediction accuracy.

You can find more information about the project here.

For my third mechanical keyboard (no, I don't have a problem), I mounted switches, soldered electrical components, and assembled a DIY kit of an ergonomic, split hand keyboard. This project was purely for fun and learning how to solder!

Check out the play by play assembly

What started as a small side project to learn SVG and the web visualization package D3.js slowly morphed into the website you're viewing right now. Every day I found another fun thing to explore, like this depiction of my past work or the outdoors visualization below, until I thought I might as well turn my standalone projects into one cohesive site all about me. I'm excited to keep working on new projects with D3.js and my abundance of personal GPS logs and Counter Strike data.

You can find more information about the outdoors visualization below here.

Speaking of my Counter Strke: Global Offensive addiction, during my year in graduate school I created a parser in Java that would read server logs from professional matches and convert the millions of lines of text into complex, usable inputs to a regression based model that predicts round by round victory probability. Perfect information of the game allowed me to incorporate precise inputs like accuracy, team economy, or utilization of support items over time into the R based model.

Here's a writeup of the project.

For my Capstone Master's Project, I designed an event based simulation for the World Health Organization alongside Professor Lee Schruben to model and optimize the high-level behavior of various worldwide staffing algorithms. We devised multiple metrics and visualizations which I wrapped in an easy to use VBA Excel interface allowing the client to rapidly test new algorithms and compare results across simulations.

Here's the final report.

During my last semester in undergrad, I led a group of five students in unifying and analyzing data across SONOS to create a dynamic, high-level dashboard. I also researched a modified Bass Diffusion model to increase demand forecasting above 90%.

Have you ever thought that boarding a plane is incredibly slow and inefficient? In my senior year, I led a group of five students to design and create an event based simulation that evaluated the overall speed and efficiency of various boarding algorithms.

Check out a presentation we made on the project.