Presented at .conf19
I spearheaded a data exploration project analyzing blackjack games in real time to predict hand-winning strategies. My team presented the project at .conf19, Splunk’s annual consumer conference with over 10,000 attendees in Las Vegas.
We used card values and win or loss markers to train a machine learning model within Splunk's MLTK
The model was retrained every 20 minutes in order to use the live conference data to continue improving
Splunking Ping Pong
Presented at .conf18
As a sales engineer intern, I collaborated with active sales engineers to build a system analyzing ping pong in real time within Splunk. We presented the project at Splunk's .conf18, at Disney World.
We collected data from multiple sources including a cell phone running custom-built computer vision software to track the ball and an inertial measurement unit from each of the ping pong paddles
I wrote calculations to extrapolate the height of the ball using the size of the ball and the x&y coordinates -- to display as a 3D scatterplot on our dashboard