Piyush
Maiti

Research

AI
Engineer

I am building AI-powered clinical assessment tools that combine speech AI, generative models, and LLM-driven pipelines to bring automated neuropsychological screening to scale, with a focus on real-world deployment in memory and aging clinics. My work includes real-time speech-to-text and text-to-speech systems, NLP scoring engines that grade cognitive test domains from voice transcripts, and computer-vision pipelines that auto-score hand-drawn visuospatial tasks using graph matching techniques. I'm focused on building clinical-grade tools that are fast, accurate, and easy to deploy, whether in the cloud or on-premise through containerized ASR/TTS infrastructure.