Built a macro signal builder that analyzes real economic indicators: GDP growth, inflation, unemployment, and interest rates, applies them across historical data for different asset classes and translates them into rule-based investment signals.

A custom-built stock screener that helps you filter and discover US equities by valuation, yield, volatility, price changes and more in a responsive user-friendly interface. Built in Python with FastAPI, PostgreSQL, and Next.JS.

TidyTable is a Python-based data cleaning automation tool that speeds up the data cleaning process for large datasets. Performs commom data preprocessings tasks like duplicates removals, missing value imputations, outliers detection, and standardizing data formats.

A minimalist web application for tracking daily habits without the friction of account creation. The app uses local storage for data persistence, allows users to maintain streaks, and ensures user privacy all in a simplified habit-tracking interface.

Developed a simple logistic regression model to predict loan default probabilities of credit card clients using a synthetic dataset of 30,000+ records. The model includes comprehensive data preprocessing, feature engineering, and handling of imbalanced classes using SMOTENC.