The creation of Neuro Veritas AI began with a question that'd been forming in my mind throughout my academic and volunteer experiences: What if we could detect cognitive decline earlier, across any language, using something like speech? This idea grew out of several intersecting passions: my fascination with the human brain, my growing experience in artificial intelligence, data science, and my deepening exposure to neurological disorders through clinical volunteering and research.

During my volunteer work at Yale New Haven Hospital and my neuroscience research with a professor at Yale, I encountered patients suffering from memory degradation and cognitive impairments. These experiences were eye-opening. I saw firsthand how many families and individuals struggled with delayed diagnoses and the emotional and financial toll that came with it. I realized that early intervention could make a significant difference in patient outcomes, yet the tools for early diagnosis, especially across diverse populations, were still limited and often inaccessible.

At the same time, I had been strengthening my technical foundation through dual enrollment in college-level data science and calculus courses, where I became particularly interested in machine learning. I began to think critically about how artificial intelligence models, especially those using natural language processing (NLP) and deep learning architectures like CNNs and LSTMs, could be applied to real-world healthcare problems. I wanted to develop a tool that went beyond linguistic boundaries, offering diagnostic support regardless of whether someone spoke English, Mandarin, Spanish, or another language.

That’s when I officially began building Neuro Veritas AI, a tech startup and research-driven platform focused on AI-powered multilingual speech analysis and logic testing to detect early signs of dementia and mild cognitive impairment (MCI). I am starting with basic speech datasets, testing how subtle changes in syntax, speech rate, pause duration, and word usage could correlate with early cognitive decline. I'm currently developing an AI model architecture combining NLP and speech detection to formulate the logic testing portion of my device. Later on, I plan to create an LSTM for temporal sequence learning and CNNs for local speech pattern detection, aiming for a model that could generalize across languages with proper training data.

Simultaneously, I created a website to present our findings and promote global accessibility. I'm designing the platform to be user-friendly, educational, and scalable, with real-time analysis potential and an international focus, offering English(for now) and more languages in the future. The platform will host resource materials to promote awareness and early intervention in the near future.

What started as an ambitious idea grounded in curiosity and compassion is now growing into a startup that bridges neuroscience, artificial intelligence, and global accessibility. My work on Neuro Veritas AI reflects everything I value: ethical innovation, interdisciplinary problem-solving, and the belief that meaningful change often starts with a single question and the courage to pursue it.