ARCHITECTURE OF A DATA-DRIVEN PLATFORM FOR FORECASTING THE FINANCIAL SUSTAINABILITY OF US SMALL BUSINESSES
Keywords:
Financial stability, US small business, bankruptcy prediction, data-driven platform, machine learning, big data, risk management, credit scoring, MLOps, US institutional environment.Abstract
This article is devoted to the development of a conceptual architecture for a data-driven platform for forecasting the financial sustainability of small businesses in the United States. The paper presents the historical evolution of bankruptcy probability assessment methods, starting with
The paper explores the range of approaches from classical financial ratio models, including the approach proposed by Edward Altman, to modern machine learning methods. The need for a shift toward integrating big data, streaming analytics, and explainable artificial intelligence tools in risk assessment is substantiated.
The institutional and regulatory environment in the United States is analyzed, including legal requirements for financial data processing and the role of government institutions such as the US Small Business Administration and the Federal Reserve System. A multi-layered platform architecture is proposed, including a data source, integration, and storage layer, an analytical layer, a service layer, and a user interface. Particular attention is paid to scalability, security, model interpretability, and the implementation of MLOps practices.
