• Developed HR Policies Chatbot using ChromaDB as a vector database and meta-llama/Meta-Llama-38BInstruct model with RAG approach, integrated into Gradio to enhance HR policy dissemination and user engagement. • Automated SOP generation from raw data files using MaziyarPanahi/Llama-3-8B-Instruct-v0.3 with Prompt Engineering, integrated into Gradio to improve operational efficiency and compliance. • Employed advanced machine learning models (XGBoost) alongside time series forecasting for predicting global freight costs in Supply Chain, launched on Streamlit to optimize logistics and reduce expenses. • Tools/Technologies: PyTorch, LangChain, transformers, Sentence-transformers, xgboost, Tensorflow, tesseract-ocr, pymupdf, chromadb, PIL, fitz

Crafted and made a compelling article for the topic "FEATURE IMPORTANCE IN MACHINE LEARNING"Performed Customer Churn prediction for a telecommunications company by classifying customers who will be churned or not . Achieved highest accuracy for Logistic Regression at 79.39%Performed detailed Exploratory Data Analysis (EDA) on ICC World Cup 2023 dataset, the analysis encompassed both batting, and bowling aspects of team and player performance, considering key metrics such as runs, wickets, strike rate, economy rate, and other relevant factors.