I built a multimodal Alzheimer’s detection system using patient voice recordings, acoustic and linguistic features, and machine learning models. Its late-fusion approach classified Normal, Mild Cognitive Impairment, and Alzheimer’s stages, achieving 87% accuracy.
For SmartPhysio, I developed a real-time exercise monitoring system with MediaPipe and integrated a PDF-based patient education chatbot using LangChain and a vector database.
I also developed a voice-controlled smart home system with Gemini API, Firebase, and ESP32 for hands-free device control. I’m a final-year Computer Engineering student at Nepal College of Information Technology, with experience building and deploying Python and React applications.
