Abstract
The explosion in usage of digital payments has led to a massive growth in personal financial transaction volumes, hence the growing demand for intelligent and intuitive solutions to manage one's finances. Current personal finance applications, unfortunately, lack personalization, automation of transaction management, and a decision support subsystem with respect to Indian financial environment. In this paper, we introduce Valora, an AI-assisted Personal Finance Management (PFM) framework, that incorporates profession-based expense categorisation, automated transaction processing, budget management, expenditure analysis and future spending predictions in a seamless package.
The framework is implemented in Python leveraging the power of Streamlit, FastAPI and SQLite. A layered modular architecture ensures the separation of UI, business logic, data processing and storage tiers. Valora churns out personalized financial insights by analyzing financial transaction patterns, estimating key financial health metrics, conducting category wise risk evaluation and forecasting expenditure based on historic spending. Valora was evaluated functionally on several user roles ranging from students to professionals, freelance and business owner by testing transaction management, budget analysis, and prediction modules. Through our experiments, we showed Valora to enhance user friendliness, automated transaction monitoring, and budget support. The proposed system offers a strong extensible framework for intelligently managing personal finance and provides potential avenue for incorporating machine learning based predictive models in the near future with mobile platform support.