Using LLMs to Hear What Students Are Really Saying

Earlier this year, I had the opportunity to lead my nonprofit, Student Echo (student-echo.org), in a collaboration with the Lake Washington School District to analyze student survey data using Large Language Models (LLMs).

With support from Dr. Tim Krieger (Director of Data and Research) and my high school principal, Ms. VanderVeer, we focused on extracting insights from open-ended responses—comments that often get overlooked because they’re hard to analyze at scale.

Our goal was simple: use LLMs to help educators better understand what students are actually saying—what they care about, where they’re struggling, and what they wish could be different.

The analysis has since been shared with district educators to help inform future improvements in the student experience. I’m excited to share the full report below, which walks through the methods, findings, and a few key takeaways from the project.

Stay tuned—more student voices coming soon.

— Andrew