Design and Development of a Hybrid Rule-Based and Controlled AI Chatbot for Digital Mental Health Services

Abstract

Mental health issues have become increasingly prevalent, highlighting the need for accessible and integrated digital psychological services. Existing systems are often fragmented — communication, information delivery, and service management are separated across different platforms, creating operational inefficiencies. This study designs and develops a web-based hybrid chatbot system for digital mental health services at Corporate Psikologi Indonesia. The proposed system applies a hybrid approach combining keyword-based detection, a structured knowledge base, and controlled artificial intelligence to produce consistent and safe responses. A consultation booking feature is built into the same platform to support both informational and service-based interactions. The system was developed using a prototype-based approach and evaluated through functional testing and a preliminary user evaluation in a controlled environment. Results show that the system supports user interaction, delivers relevant mental health information, and facilitates consultation booking and service access effectively. The proposed hybrid chatbot architecture addresses service accessibility while maintaining response reliability. Future work will focus on broader user evaluation and stronger natural language understanding capabilities.