CSUN Students Win Elite NSF Fellowships Using AI for Good

Chemistry student Sammie Ellis and astrophysics major Jesus Poblano are among just 2,500 winners nationwide selected from nearly 14,000 applicants for the National Science Foundation Graduate Research Fellowship. Their secret weapon? Teaching computers to think like scientists.

Ellis is training artificial intelligence to hunt for superhydrides, hydrogen-rich materials that could make MRI machines and high-speed trains faster and more energy efficient. Traditional methods take weeks or months to screen potential materials. Her AI approach could shrink that timeline dramatically by learning to predict which candidates show the most promise.

"If we can find a way to train AI to quickly screen all these candidates, then you can narrow down the search space," Ellis explained. The breakthrough could accelerate discoveries that typically crawl along at a glacial pace.

Meanwhile, Poblano is pointing AI at the stars. He's developing machine learning algorithms to analyze light from telescopes like the James Webb Space Telescope, searching for atmospheric signatures that could indicate life on distant planets. His method identifies gases in planetary atmospheres by measuring how they interact with light.

The efficiency gains matter because telescope time is precious. "In those 50 hours, you can get more observations of other planets rather than just focusing on one," Poblano said. He's building tools that future researchers can use to confidently identify signals "that can only be explained by life, rather than a natural phenomenon."

Neither student took a straight path to this moment. Ellis started as a pre-med biology major, switched to chemistry at UC Irvine, and bounced through atmospheric and organic chemistry before discovering her passion for computational work at CSUN. Poblano joined a biophotonics lab and explored widely before finding his focus through independent study.

Why This Inspires

Both students credit their success to embracing uncertainty and trying new things. Ellis puts it bluntly: "Stupidity is one of the most important parts of research. You're looking at things you don't understand and trying to figure it out."

Their wins show how AI isn't replacing human curiosity but amplifying it. By teaching machines to handle the heavy computational lifting, these young scientists are freeing up time and resources to ask bigger questions and pursue bolder answers.

Ellis heads to Oak Ridge National Laboratory this summer before applying to PhD programs in computational chemistry. Poblano starts his PhD at Johns Hopkins University in the fall, where he'll continue developing his search for life beyond Earth.

Two first-generation students, armed with algorithms and ambition, are proof that the next generation of scientific breakthroughs is in very good hands.