About Me
At Sandia, I primarily worked on anomaly detection methods. One of the main projects I worked on was building failure forecasting models in R using Isolation Forests for lithium-ion batteries that were cycled to failure (defined as a certain percentage of the nominal capacity) under different fast-charging conditions. Other projects I worked on were for developing anomaly detection methods for acoustic signals and images in Python using TensorFlow. I also provided an advisory role for other engineering summer interns on developing anomaly detection methods for acoustic signals from a statistical perspective.
During my internship, I worked on two separate projects that made use of the firm's proprietary Functional Information System (FIS). In the first project, I developed classification models using algorithms typical for natural language processing in order to classify job postings to its appropriate FIS code. In the second project, I analyzed geographically proximate communities for patterns in functional distributions. This work was ultimately presented to the entire company and investors.
Opticlose is a startup aiming to provide high volume sales teams the tools and technology to more effeciently make sales. I specifically analyzed sales data using a combination of Excel and R. I created models to predict the success of a sales closure at each step of the process. In addition, I created a company specific library of R scripts to better automate the process in the future. I also created and presented a slide deck consisting of my analytical results to investors.