I am Sumukh, an extremely passionate Machine Learning/Software Engineer with over 2.8 years of experience.
I’ve previously worked as a Machine Learning Engineer, building and integrating Generative AI-powered product components using tools like Python, React, Computer Vision and Large Language Models, and more. I’ve contributed to both fast-paced startups and innovation-driven teams at large companies, helping build novel, high-impact AI solutions.
Currently I'm strengthening my fundamental knowledge at San Jose State University by pursuing a Master’s in Computer Science.
Ex-ML Engineer @ Inhabitr | Software Engineer @ Zebra Technologies
Interests: Generative AI, LLMs, ML Systems, Full-Stack development
🧪Current Research: Enhancing code review automation using enhancements on Quantized 'Dynamic' Low-Rank Adaptation (QDyLoRA) fine-tuning with Llama 3 and integrating a RAG pipeline for better inference, focusing on performance-efficient open-source solutions. Goal is to reduce memory utilization during fine-tuning while making sure inference quality is not compromised.
Thank you so much for stopping by! I am always up for collaboration or conversation — drop a star, fork, or a message!
