Here I build and maintain Machine Learning Systems, Deep Learning Architectures and Open-Source package.

• 🔭   I'm currently maintaining Upasak
• 🌱   I’m currently learning MLOps, Mathematics and Generative AI
• 😵   Buried under backlog of papers to read.

Hi, I’m Shrut 👋

I am a firm believer of figuring out as we go.

Learn - Build - Feedback - Iterate

I'm a Machine Learning Engineer with a year of experience working on applied ML problems in MedTech, while navigating through ambiguity and vague ideas.

I started my career in 2024. And started exploring AI at more deeper level to go beyond thinking it as black box, and as I learned more about AI, I fell for fundamental mechanism and principles of Deep Learning architectures.

I'm broadly interested in maths, finance, tech, and ways to stop self-sabotaging. Aiming for generalist status, but my comfort zone has me in a death grip. Don't believe in fluff projects, either you learn or build something that solves problem.

All of my projects are released as open-source on GitHub:

• upasak - A flexible, mindful to privacy, no-code/low-code framework for fine-tuning large language models, built around Hugging Face Transformers. It offers an easy-to-use Streamlit-based interface, multi-format dataset support, built-in PII and sensitive information sanitization, and a customizable training process.

• attention_to_llm - My journal for hands-on exploration of building and training large language model from scratch.

• text_recognition - An end-to-end OCR pipeline with a web-based upload interface using CRAFT for text detection and CRNN for text recognition. Models were trained from scratch on ~45,000 images, with experiments and versions tracked using CometML.

• retail_vision - A pipeline designed to detect and group products on retail shelves. This project built using flask utilizes multiple microservices to process images, detect products using a YOLO model, and group the detected products by category.