Financial Charting
Working for 6800 Capital LLC, a
hedge fund in Princeton, NJ, I developed Independence, a
financial charting and strategy back-testing application for commodity
trading. The application, written in Visual Basic, downloaded and
cleansed financial data and entered it into a fully normalized database.
Portfolio Optimization
Consulting for Rogue Wave Software, I developed a program to
visualize portfolio volatility and generate trading signals in commodity
markets for Global Advisors, a hedge fund based in London. The process
to test the signals was run nightly to adapt to changing market
conditions and used both linear and nonlinear optimization routines.
This process was then highly parallelized to reduce the amount of time
it took to run from hours to minutes. Visualization of the trading
signals allowed for human verification of the optimized results.
Visualization techniques included 3D surfaces and contours.
Flight Path Simulation
Agusta Westland routinely generates massive data sets
running Monte Carlo type simulations on helicopter flight paths in order
to minimize helicopter downtime and optimize replacement part
shipments. These large datasets were difficult to understand and
error-check. Consulting for Rogue Wave Software, I used the Visualization
Toolkit (VTK) to develop a 3D visualization playback tool to display
these files. This project was managed using the agile software
development model.
PyIMSL Studio Demos
Rogue Wave Software launched a new visual data analysis
framework centered around Python and Matplotlib in late 2008. I was
solely responsible for developing the demonstration suite to showcase
this new technology. The demo suite was cross-platform compatible and
consisted of demonstrations of the Neural Network, ARIMA and K-Means
algorithms. It was written in WxPython and used the Matplotlib graphing
framework. 
PV-WAVE
One of my primary responsibilities working at Rogue Wave Software is
the support of the PV-WAVE programming language. Several notable
upgrades to the language included the addition of MySQL database
connectivity, addition of Neural Network routines and an improved
demonstration gallery.
