🇺🇸Kavya
Kopparapu

Kavya Kopparapu was sixteen years old and looking at a problem that most medical researchers had spent careers failing to solve when she decided to apply machine learning to brain tumor diagnosis. The specific challenge was glioma grading — the classification of malignant brain tumors into grades that determine treatment protocols and prognosis. Grade II gliomas are slow-growing and often managed conservatively; Grade III and IV tumors are aggressive, with median survivals measured in months. Accurately grading a glioma requires analyzing tissue samples at the molecular level: identifying specific gene mutations like IDH status and EGFR amplification that cannot be seen with conventional pathological staining. The standard method required expensive genetic sequencing that many hospitals, especially in the developing world, could not afford.

Kopparapu built GlioVision: a deep learning system that analyzed histopathological images — microscope slides of tumor tissue stained with standard dyes — and predicted molecular tumor grade with high accuracy, without genetic sequencing. She trained a convolutional neural network on thousands of labeled pathology images, teaching the model to recognize the subtle visual patterns that correlate with molecular markers. The system achieved accuracy rates that compared favorably to specialist-level diagnosis on the datasets she tested, and it required only a standard microscope slide rather than expensive sequencing equipment.

The implications were concrete. In high-income healthcare settings, GlioVision could accelerate diagnosis and reduce costs. In lower-resource settings — rural hospitals in India or sub-Saharan Africa — it could make brain tumor grading accessible where genetic sequencing was simply not available. Kopparapu had built a tool that democratized a diagnostic capability: accurate molecular glioma grading, delivered through image analysis, available anywhere a microscope and a laptop could reach.

Alongside her medical AI work, Kopparapu had already founded Girls Computing League at sixteen — a nonprofit organization dedicated to increasing access to computer science education for girls, particularly those from underrepresented backgrounds. The league ran workshops, provided mentorship, and built curriculum specifically designed to engage young women who had not previously seen themselves as belonging in technology. By the time Kopparapu was featured in Forbes 30 Under 30, GCL had reached thousands of students and operated chapters across multiple states.

The dual track of her work — clinical AI and equity advocacy — reflected a coherent philosophy. Kopparapu understood that building powerful AI tools was only half the project; the other half was ensuring that those tools were built by people from all backgrounds, and that they reached all populations. GlioVision addressed access in diagnosis. Girls Computing League addressed access in creation. Both interventions targeted the same underlying problem: the unequal distribution of technical capability across geography, gender, and economic circumstance.

Kopparapu enrolled at Harvard University, where she continued pursuing research at the intersection of artificial intelligence and healthcare. Her Congressional App Challenge win — a competition for student app developers — added to a portfolio of recognition that included national science competitions, media features, and invitations to speak at conferences attended by professional researchers. She was not yet twenty years old when she had built a cancer diagnostic AI, led a national nonprofit, and landed on one of America's most recognized young innovator lists. The brain tumor patients whose diagnoses her tool could improve were not abstractions to her — her grandfather had suffered from a glioma, and the inadequacy of available diagnostic tools had been personal before it was intellectual.