Time series solution for evaluating portfolio Value at Risk (VAR) with a deep learning model, resulting in a 10% reduction in estimation time and better investment decisions for clients.
Tech stack: Python, Web Scrapping, MySQL, plotly, LSTM, RNN, Pyfolio, TensorFlow, NumPy, Pandas, Flask, and Scikit-learn
Service implemented in MLflow model using Python to develop a human activity recognizer that interprets human body gestures via sensors. Accurately determine and classify human actions and activities.
Tech stack: Python, TensorFlow, NumPy, Pandas, Flask, MLflow, Streamlit, AWS, and Scikit-learn