When I was nine, my grandmother in Taipei went to the hospital for a persistent cough. She was seen within twenty minutes, given a chest X-ray and a diagnosis, and walked out with antibiotics — total cost, about six dollars. Four years later, after my family moved to California, my mother waited three weeks for a primary care appointment for a cough that turned out to be the start of pneumonia. I remember sitting in the waiting room doing the math on the bill before the doctor had even walked in. Same symptom, same family, two systems that treated urgency completely differently. That gap is where my interest in science actually started — not in a textbook, but in the question of why two countries with similar wealth and technology could deliver such different outcomes to the same kind of patient. That question pulled me toward neuroscience and data, because I wanted to understand health problems at the level where they're actually solvable — in the signal, not just the symptom. Through the Intelliscience Institute, I've spent the last two years building machine learning models that detect epileptic seizures from EEG data, working with researchers at San José State University. I built a full pipeline in Python — preprocessing raw EEG with MNE, then training an ensemble of SVM, Random Forest, and XGBoost models — that reached 93.6% accuracy on the CHB-MIT dataset, work I later presented as a poster at NCUR 2026 and co-authored into an IEEE survey paper on AI-driven EEG brain health analysis. What fascinates me most isn't the accuracy number, but what it represents: a tool that could someday make seizure detection as fast and cheap as my grandmother's diagnosis was. Good science, to me, is fundamentally about closing gaps like the one I watched my own family fall into. But research alone doesn't close gaps — it has to reach people who can act on it loudly enough that something changes, which is where advocacy comes in. I founded Bay Area NutriLens after noticing a smaller version of the same problem: students at my school eating lunches that technically met state nutrition guidelines but were, by any nutritional measure, poor. I built NutriLens into a district-wide research effort — collecting menus, cross-referencing them against USDA standards, and putting findings directly in front of people with the power to act, learning a skill research alone never taught me: translating data into language a school board, not a science fair judge, would act on. I've carried that instinct into Bay Area Medical Explorers, where I serve as the social media manager, run weekly meetings with our core team, build our website, and write the social media content that reaches students in the first place — because a resource nobody can find isn't accessible at all. Science doesn't defend itself. It has to be carried, by someone willing to sit through a budget meeting or post when the engagement is silent, and explain, slowly and clearly, why this matters. What's at stake when science isn't championed is exactly the gap I started with — not abstract scientific literacy, but real people waiting three weeks for a diagnosis that should take twenty minutes, or eating lunches that meet a checklist instead of a standard. Science doesn't fail because the discoveries aren't there. It fails when nobody pushes them far enough to reach the people who need them. That's the advocate I'm trying to become: not just someone who runs the analysis, but someone who stays in the room until it changes something.