Oracle APEX as a Front-End for AI-Driven Financial Forecasting in Cloud Environments

Abstract

This paper presents a cloud-native architecture that integrates Oracle Application Express (APEX) with Python-based deep-learning models for AI-driven financial forecasting in Oracle Cloud Infrastructure (OCI). The framework employs LSTM and GRU networks trained in TensorFlow/Keras to model nonlinear temporal dependencies in financial time series, while APEX serves as the interactive front-end for visualization and user control. RESTful APIs connect APEX dashboards to OCI Functions hosting the model containers, enabling real-time forecast generation and dynamic result rendering with minimal latency. Experimental evaluation demonstrates high predictive accuracy (R² = 0.972), low mean latency (142 ms), and strong cost scalability across increasing user loads. The study further discusses implications for enterprise deployment, highlighting security, scalability, and maintainability considerations essential for financial applications. The proposed approach establishes a reproducible foundation for embedding adaptive AI forecasting pipelines within low-code environments, paving the way for future integrations involving AutoML, ONNX-based model interoperability, and self-updating financial forecasting systems.