Aspiring Data Scientist with strong Python skills and experience in building and deploying ML models. I am familiar with LLMs, RAG systems (LangChain, LlamaIndex), and vector databases like Pinecone.

Skilled in NLP, Computer Vision, and Deep Learning using tools like TensorFlow, PyTorch, Hugging Face, and OpenCV. Proficient in data manipulation with Pandas and NumPy, and integrating ML models with APIs for real-world applications.

Machine Learning & AI: Computer Vision, NLP, Deep Learning, Machine Learning, LLMs, RAG

Frameworks & Libraries: TensorFlow, OpenCV, YOLO, scikit-learn, Pandas, NumPy

Language: Python, SQL

Development Tools: Git, FastAPI, Flask, Docker, Vector Databases

• IntelliShelf-Store: IntelliShelf is a full-stack AI-powered platform that combines Computer Vision, LLMs, and RAG to create a smart e-commerce experience for musical instruments. It offers instrument classification, YOLOv8-based defect detection, LLM-generated descriptions, and a LangChain-powered chatbot for FAQs and returns. Built with FastAPI and Docker (ARM64), it's ready for deployment on AWS, Oracle, and Azure.

• Rag-Assistant:A production-ready RAG (Retrieval-Augmented Generation) web application that leverages Google Gemini Pro for question answering using user-uploaded PDFs or CSVs. Built with FastAPI, LangChain, Pinecone, and LangSmith for robust LLMOps, and deployed using Docker, GitHub Actions, and Render.

• Finetuning-on-OpenSource-LLMs-using-Cleaver-Techniques-like-LoRA-and-QLoRA: Fine-tuned on a custom dataset to generate domain-specific responses, as well as on an instruction-based dataset for enhanced question-answering capabilities.

• Automate Stocks and Cryptocurrencies Research: Automates scraping, summarizing, and sentiment analysis of financial news using Python and Hugging Face models to streamline stock and crypto research.

• Drowsiness Detection: A Deep Learning Custom Model that detects driver drowsiness in real-time using a custom YOLOv5 model.

• 📫 How to reach me yashh2417@gmail.com

• Automate Stock and Crypto Research Using Python and Machine Learning
  Published in: 2025 3rd International Conference on Disruptive Technologies (ICDT) · May 13, 2025
  Description: Real-time summarization and sentiment analysis of stocks and cryptocurrencies to support forecasting and investment insights.