AstraGenX User

I'm a Computer Science undergraduate building full-stack AI products and researching safer clinical language models. At Jadavpur University, I benchmarked Pegasus against quantized LLMs including Llama-3, Qwen, and BioMistral for clinical text summarization, identifying a 38% hallucination rate in modern LLMs versus 4.35% in legacy architectures.

I've modernized aerospace software workflows at Hindustan Aeronautics Limited and built products including AstraGenX's startup platform and a Medical AI Benchmarking Dashboard. My work spans React, Node.js, Python, NLP, quantized LLM inference, and deployment-focused AI evaluation.