Portal 02
A few things I’ve built while exploring AI, machine learning, data, and software. More experiments and ideas will find their way here.
A machine learning system that uses customer transaction history to estimate a digital trust score based on user behaviour. It combines feature engineering, XGBoost, model calibration and SHAP explanations, with a FastAPI service and web interface for testing different customer scenarios.
A healthcare navigation app that helps users understand possible conditions from their symptoms and assess diabetes risk from basic health and lab values. It uses machine learning to rank possible conditions, explain risk factors and suggest relevant tests and specialists.
A Python based platform for analyzing stock market data, risk and trading strategies. It covers technical indicators, statistical analysis, volatility forecasting, market regime detection, predictive models and backtesting to evaluate ideas against historical data.
An Android application built to bring common college activities into one place. Students can access notices, study material, attendance and syllabus progress, while teachers and staff can manage attendance, files and academic updates.
