kaggl

A Flask-based ML application that predicts blood groups using fingerprint images. It integrates a TensorFlow (Keras) model with 89% accuracy, featuring user authentication, database management with Flask SQLAlchemy & SQLite, and a frontend built using flask. 🚀

Solution of the Titanic Kaggle competition: Demonstrates basic data munging, analysis, and visualization techniques. Shows examples of supervised machine learning techniques.

End-to-end Spotify data analysis using Python (Pandas) and Tableau, uncovering artist, genre, and track popularity trends through data cleaning, preprocessing, and interactive dashboards.

Kaggle competition

Production-ready NLP pipeline for Multilingual Natural Language Inference (NLI). Built for the Kaggle "Contradictory, My Dear Watson" competition using PyTorch Lightning, LoRA, and MLflow.