Implementation of a Multi-Platform Hybrid Chatbot for Digital Customer Service (Case Study: Batik Wistara, Surabaya)
Abstrak
Micro, small, and medium-sized businesses (MSMEs) have been motivated to provide quicker, more responsive, and consistently available customer care due to the swift development of artificial intelligence (AI) and digital transformation in the e-commerce ecosystem. Due to consumer desire for comprehensive product information, such as motifs, color variants, price, stock availability, and purchasing methods, real-time engagement is becoming more and more crucial in traditional batik enterprises. However, limited working hours, service capacity, and information consistency limit customer support systems that depend on human operators, which frequently results in delayed responses and lower customer satisfaction. This project integrates rule-based mechanisms, Natural Language Processing (NLP) with synonym-based token expansion, and Groq AI as a Large Language Model (LLM)-based conversational reasoning engine to create a hybrid chatbot system for UMKM Batik Wistara in order to address these issues. Using a two-way webhook method, the chatbot functions as a multi-channel customer support solution that is implemented throughout the Batik Wistara website, the WhatsApp Bot API, and the Telegram Bot API. For traditional batik MSMEs, the system's complete integration with the Batik Wistara e-commerce database enables precise, real-time information distribution and increased service efficiency. In addition, the chatbot supports automated responses that ensure consistent information delivery across all platforms. This implementation is expected to enhance customer experience while reducing the operational workload of business owners.
Identifier
- DOI: https://doi.org/10.63158/scd.v3i2.55
- OpenAlex: https://openalex.org/W7171302873