Internship : FinvestFx
At FinvestFx, I gained in depth knowledge about the FX market in India and current FX processes. I learnt how the best rates for FX can be provided to the customer thereby reducing the cost and managing FX risk. I understood the basic risk management processes that need to be followed to manage FX risk for an organization. I worked to improve their website and communication material, enhancing efficiency and functionality
My primary contribution was developing the FX Data Forecasting project that is now being offered as an extra service to clients of Finvestfx. The project is a financial data analysis and predictive modeling system. It involves analyzing daily USD-EUR exchange rate data to generate accurate forecasts for the next five days. The process begins with the careful preparation of the data, followed by the generation of important financial indicators that help capture market trends and behavior. Two forecasting models are used: a machine learning-based model (LightGBM) and a time-series forecasting model (Prophet). The purpose of these models is to predict the closing prices for the next five days, using historical exchange rate data.
​
The data preparation process ensures the accuracy of the dataset, with any missing or incomplete values being handled to maintain consistency. Financial indicators such as market momentum and price trends are generated, providing critical inputs for the machine learning model. The LightGBM model utilizes these indicators to make informed predictions about future exchange rates. Additionally, the Prophet model focuses solely on time-based patterns to offer predictions. Both models' accuracy is measured using a performance metric that compares predictions with actual exchange rates, ensuring reliability.
​
This project highlights my ability to analyze and forecast financial data, making use of advanced models to predict future market behavior. By successfully applying machine learning and time-series forecasting, I was able to demonstrate my understanding of financial analysis techniques and predictive modeling.
