Name: Erik Palmer
Pronouns: he/him/his
Biography:
My name is Erik Palmer and I will be the point of contact for this co-mentored project. I am a Software Integration Engineer for the National Energy Research Scientific Computing Center (NERSC) at the Lawrence Berkeley National Laboratory. I joined the User Engagement Group to support software installation and continuous integration services. For my Applied Mathematics Ph.D. I developed GPU code to simulate the behavior of polymer gels. In my past, I spent 4 years at community college, was an underwriter for car insurance, English teacher in Beijing, Math instructor and umployed pandemic stay-at-home dad. I work mostly remote from East Lansing, Michigan. Lipi Gupta is a Sceince Engagement Engineer at NERSC with the User Engagement Team at the Lawrence Berkeley National Lab. At NERSC, she helps users engage with staff and with each other to bridge gaps in scientific computing. She obtained her Ph.D. in Physics with a background in nonlinear beam dynamics. She conducted research at the SLAC National Accelerator Lab, where she studied the application of machine learning techniques for improved particle accelerator operation and control. Lipi enjoys STEM related outreach and endeavors to promote young women and minorities to pursue science careers.
Institution/Lab: Lawrence Berkeley National Laboratory
Website: https://www.nersc.gov/about/nersc-staff/user-engagement/erik-palmer/
SRP Collaboration Topic/Title: Supercomputer User Ticket Modeling and Analysis for Improving Scientific Output
Field or research area: Data Science/Applied Math/HPC
Please select all the topical areas that apply to your project:
Data Science (i.e., data analytics, data management & storage systems, visualization); High-Performance Computing; Machine Learning and AI
Brief Abstract:
At the National Energy Research Scientific Computing Center (NERSC), we cater to a user base exceeding 9,000 individuals, facilitating groundbreaking scientific endeavors on the 8th fastest supercomputer globally. As part of our support team, our primary mission is to assist scientists and researchers in using this powerful resource for a variety of scientific computation. Our responsibilities encompass troubleshooting application issues, optimizing performance, and maintaining the computing hardware and software environment. In order to do this, we would like to make data-informed decisions based on user behavior by analyzing support ticket content to identify and address user needs strategically. For this project, we would like to use our ticket text information to gain insight into user behavior, sentiments and needs regarding high performance computing. We are open to working with a variety of tools and techniques to meet this goal. This could include machine learning driven analysis techniques such as natural language processing and data clustering powered by our own supercomputer. Or it could involve simple statistical analysis techniques as these models can also be effective tools for aiding understanding. A successful project will produce data artifacts that help identify user trends or behaviors to guide decisions about user services.
Desired relevant skills, background, or interests:
We would like students who value listening skills, exhibit perseverance by working alone or reaching out for help from others. Students who regularly and proactively check-in for assistance and progress updates are also highly desired. Knowledge of coding, statistics, machine learning, and language processing are useful, but not required as mentees will be given time, resources and our guidance to learn. The project mentors would like to use Python as the primary language, so familiarity with Python is ideal. We believe there is a lot of room to move in different directions with this project depending on your own interests.
Other comments:
Do any special requirements apply? Permanent Resident OK; International OK
Other, specify:
Keywords:
high performance computing; nlp; natural language processing; ml; machine learning; python; data science; statistics; modeling; graph theory; data analytics
Lightning Talk Title: Supercomputer User Ticket Analysis for Improving Scientific Output