Two Weeks, One Challenge, Lifelong Impact: Bobcats Dive into Data Science Challenge

Kathy Chau believed she knew what her future would look like. The first in her family to attend college, she had long been advised to aim for a safe and stable job - one that might not light a fire in her soul but would pay the bills.

"I resigned myself to working a corporate job. I didn't like the sound of it, but I didn't hate it, either," she said.

That was before she spent two weeks immersed in the Data Science Challenge at Lawrence Livermore National Laboratory, a hands-on program that brings together undergrads, graduate students and lab mentors to solve a real-world data science problem.

Chau experienced a turning point in thinking about her future, she said, when LLNL Director Kimberly Budil visited the participants and told them how scientists at the laboratory's National Ignition Facility had achieved fusion ignition after decades of trial, error and scientific grit.

"Her account echoed a passion I had seen in each of the staff members I met," the Hayward native said. "They had this unity in pushing the boundaries and working toward a common goal, and it struck me that I wanted to be somewhere where I, too, could speak about my job with such passion and excitement."

Suddenly, her plan for a safe, stable job in some nameless corporation seemed even less appealing.

"I knew then that I wanted to work at the lab," Chau said.

Training, Tours and Takeaways

This year, the program's sixth, 43 students attended the program from UC Merced, UC Riverside, UC Santa Cruz and California State University, Bakersfield, the largest number of participants and schools yet.

This year's challenge focused on using computer vision techniques to analyze synthetic data - training models that could support automated laboratory work, such as controlling robotic arms or processing hazardous samples. At the end of the two weeks, each team presented its findings at a scientific poster session.

Days at the DSC are packed. Students dive into the Challenge problem, attend professional development sessions and hear from scientist-mentors. They tour high-tech facilities, including the jaw-dropping National Ignition Facility, where researchers push matter to pressures and temperatures that make the sun's core look tame.

For many students, the DSC is their first exposure to the world of research beyond the university, as well as internship opportunities and possible careers at national labs. Lab mentors are carefully selected to represent different career stages, from postdoctoral researchers to senior staff, so they can discuss their career paths with participants.

"My goal is to share tangible, real-world information about careers that might not be well-known, especially those related to national labs," said UC Merced applied math Professor Suzanne Sindi, a DSC co-founder. "Career development is a major part of this; it's about professional growth and learning. For me, success is measured through the career trajectories of individual participants."

Sindi's passion is deeply rooted in her own life journey.

"I wasn't the first in my family to go to college, but I was the first to go into a STEM field. I loved problem-solving and majored in math," Sindi said.

That led her to undergraduate research and eventually to graduate school, where, late in her studies, she learned that many of her peers had completed internships at places like the Jet Propulsion Laboratory near her Southern California home.

"I didn't know you could do an internship there," she said. "I'm happy in academia, but I think about all the other doors a STEM degree can open - if students know where to look."

From Nerves to Networking

Hosted just outside the LLNL gates at the University of California Livermore Collaboration Center, the DSC aims to "remove barriers and set up students with the skills and training employers want, and give them as many options as possible," Sindi said.

Sometimes the barriers aren't about knowledge of opportunities, they are about nerves. Working with professional scientists can feel intimidating at first, participants said. But everyone agreed that the staff members at LLNL were warm, welcoming and always available for help or to talk.

Tahira Williams, a fourth-year graduate student originally from Kingston, Jamaica, applied to be a team lead for the 2024 Data Science Challenge at the urging of a friend and fellow graduate student.

"At first, I was nervous," she said. "Most of my machine learning background is based on experience, not formal classes."

She leaned into what she knew: biology, technical presentations and science communication, and was accepted as a team lead, guiding three undergraduates through a project using machine learning to predict heart health based on ECG data.

Challenges? There were plenty, especially when it came to finding enough data to train their machine learning model. Also, team members - last year, the undergraduates came from UC Merced, UC Riverside and Case Western Reserve with varying levels of experience - had never worked together before. But she said the team divided the work according to each member's strengths and took the lead on their respective parts of the project and final presentation.

"It was great to see them empowered and confident," Williams said.

Moitrish Majumdar, a third-year Ph.D. student in applied mathematics, also served as a team lead last year. When another team lead didn't show up, Majumdar's small group merged with it, doubling its size.

"The team was a mixed bag," they said. "Some had a lot of programming experience; some didn't. It took time to understand the dynamics, but it helped me learn about managing and leading, which are two very different things."

Majumdar said they appreciated the availability of help from the organizers and other team leads. They also said the DSC's networking opportunities helped them overcome an initial shyness.

"We were encouraged to talk to everyone and ask a lot of questions," Majumdar said. "It was sort of like exposure therapy - intimidating at first, but very effective."

For many students, the DSC is their first chance to work on a real-world research problem. It's also their first time interacting daily with professional scientists-and sometimes realizing those scientists aren't so different from them. Everyone who participated stressed the importance of taking advantage of networking opportunities, not only with the professionals but with peers who, as Chau said, have their own experiences and insights.

Majumdar and Williams both landed internships this summer: Williams at LLNL as a Computing Scholar Fellow, and Majumdar at the Lawrence Berkeley National Laboratory. Both credit the DSC with helping them get there by networking and boosting their resumes, and plan to apply for more internships and fellowships.

Williams' advice: Talk to the scientists. Ask about internships. Be honest about what you're hoping to learn and where you want to go.

"When I applied for my internship, I reached back out to scientists I'd met during the Challenge," she said. "I asked if they remembered me, and they did."

That network - and the self-assurance that came with it - has stayed with her.

"You get to do something you're not confident in, and that's where you experience the most growth," she said. "The DSC is one of those experiences that will be with me forever. I came in just knowing the basics, and I left being able to communicate with people who have been doing this for years."

A Launchpad for Careers - and Friendships

Ryan Milstrey, a spring graduate from Folsom, joined this year's Challenge to build his technical experience in deep learning for computer vision. Chau wanted to experience working in a high-pressure situation. Majumdar and Williams wanted the growth and networking experience.

They all got more than they expected.

"The key thing I gained from this experience was the connections I made - with the DSC staff, with my peers, and with the scientists at the lab," Chau said. "Brian (Gallagher, DSC co-director) told us not to overlook the value of our peer network. Thanks to conversations with other students, I learned about research programs I hadn't heard about before, such as URISE and CITRIS, and the staff gave us tips on preparing resumes and applying for the lab's summer internships."

Majumdar, who studies computational and mathematical biology, said, "The DSC gave me the opportunity to apply related skills in a different context."

"I learned so much, and I got a chance to talk about science that I wouldn't normally talk about in my own thesis work," Williams said. "The DSC was a great leadership development opportunity."

"I expected people would be secretive about their ideas, but they were very open and the discussion was valuable and interesting," Milstrey said. "Competition between the teams was minimal. It is a really low-stress learning environment."

It's also a collegial environment. The students not only work together all day but stay at the same hotel, share meals and spend free time together.

Some students visited Yosemite National Park for the first time, while others explored the Bay Area. Williams and Majumdar, who remain close friends, took a weekend trip to Point Reyes National Seashore during their DSC.

"It was so beautiful," she said. "We saw whales in the distance - I had never seen a whale before. It was just wonderful."

"I've never been to summer camp, but to me it feels like summer camp because everybody starts off nervous and just getting to know each other," Sindi said. "And then by the end of this intense time working together, there's a camaraderie that I hope is carried forward with them."

Not everyone had the career epiphany that Chau did. For Milstrey, the DSC cemented his interest in pursuing computer vision applications of deep learning research as a career.

The others are still considering their options.

"The organizers introduced me to scientists who had biology backgrounds and shifted into computational work," Williams said. "That helped me see a path for myself." She hasn't decided on a career path yet, but is sure she wants to use data science to improve health outcomes.

"The DSC opened up my horizons, showing me what it's like to work at the lab and informing me about the different types of roles that exist within it, as well as how the lab is very collaborative," Majumdar said. "I haven't decided what I want to do, but the DSC has helped me decide that I want to work in a national lab or industry."

The one thing they all agreed on was that anyone who's even just a little bit curious about data science or working at a national lab should apply for the DSC, even if they don't think their skills or resume match 100%. Organizers are already working on plans for next year's challenge, and applications will be available on the program's website.

"Don't be afraid. You're there to learn and grow," Williams said. "Leverage the opportunity you're given."

Lorena Anderson