Real-time causal ads optimization: an integrated web-based data collection and automated analysis pipeline

This project focuses on designing and implementing a web-based application for real-time advertisement display and interaction logging. The platform, built using Java and Spring Boot, captures ad impressions, clicks, and contextual metadata, which are stored in a structured database. In Part 2, causal machine learning techniques are used to estimate the uplift (incremental effect) of ads on conversion and engagement. By integrating internal ad logs with external datasets (e.g., Criteo Attribution Dataset), the system enables real-time automated ad targeting and optimization.