<p>Hypertension remains a leading cause of cardiovascular morbidity and mortality worldwide, highlighting the need for continuous and non-invasive blood pressure (BP) monitoring. Photoplethysmography (PPG) sensors embedded in wearable devices provide a promising pathway for cuffless BP estimation. While several reviews have summarized PPG-based BP monitoring, existing studies often focus on either sensor technologies or AI models in isolation. This review differentiates itself by presenting an integrated perspective on PPG-based cuffless BP monitoring that spans from signal acquisition and preprocessing to advanced artificial intelligence (AI) algorithms and their real-time deployment on edge devices. The main contributions include: (i) a systematic synthesis of PPG sensor technologies, datasets, and signal processing methodologies; (ii) a comprehensive comparison of machine learning and deep learning approaches for BP estimation and classification; and (iii) an in-depth discussion of edge-AI frameworks, highlighting model optimization, quantization, and deployment on platforms such as STM32 microcontrollers and Raspberry Pi. We further identify existing challenges such as motion artifact suppression, dataset limitations, and model generalization issues, and provide recommendations for overcoming these barriers. By consolidating both the theoretical foundations and the practical edge-computing implementations, this review provides a unique roadmap for researchers and practitioners toward developing clinically viable, low-power, and real-time wearable BP monitoring systems.</p>