Varadkrishna
Kamtikar

End-to-end, leakage-free ML pipeline detecting reward-related neural patterns in EEG signals across 32 subjects from the DEAP dataset. Rigorous cross-subject cross-validation, SMOTE oversampling, and an interactive Streamlit dashboard.

LSTM-based stock price prediction dashboard for Indian equity markets (NSE & BSE). Forecasts prices from historical time-series data using deep learning with an interactive visualisation interface.

Real-time weather analytics platform for Ireland. Integrates Python, PostgreSQL, and XGBoost to forecast rainfall, with Streamlit dashboards and live data ingestion pipelines.

Interactive platform visualising Ireland's rental and property market using RTB & CSO government data. Tracks rent trends, affordability, and regional patterns across all Irish counties with Plotly dashboards.

ML pipeline to detect fraudulent credit card transactions. Applied classification algorithms with feature engineering, SMOTE for class imbalance handling, and rigorous precision/recall evaluation.

ML-powered REST API using FastAPI to classify bank notes as authentic or fake in real time. Scikit-learn model served over Uvicorn with four image-derived features as input.

Tableau dashboard examining Netflix catalogue patterns across content type, geography, ratings, and duration. Interactive visualisations reveal genre trends and global distribution across the streaming library.