Software projects of past and present

Realtime visualization of City Energy Usage
Utilized statistical models, along with public data to create a beautiful visualization of Manhattan’s energy usage. I wrote a framework called pyshader to render 8K resolution video directly from OpenGL

A company I co-founded that built business intelligence software for Hadoop. The company raised $2.5mm in funding from investors such as Accel Partners, Amr Awadallah (CTO Cloudera) and others. The company was acquired in 2014.

Modeling Vehicle Behavior for Autonomous Driving
I worked on Stanford’s DARPA Urban Challenge entry “Junior”. Our project experimented with improving the monte-carlo localization algorithm that precisely positioned Junior to centimeter level accuracy. I also worked on building statistical models to predict the motion of near-by vehicles from LIDAR scans with the hope of improving safety during lane changing.

Motion Planning for Skateboard-like Robots in Dynamic Environments
In this article, I present new techniques that better compute paths for vehicular robots, specifically those that move through terrain that affects the dynamics of vehicle motion to a large degree (for example: a skateboarder in a skatepark or a jeep driving over rocky, inclined terrain)

Patellar Ligament Tear Detection using Machine Learning
In this paper we show a method for identifying ligament tears in MRI scans. We present a new algorithm that builds a spatial pyramid over the tendon invariant to scale and orientation changes of the tendon.