Nice Intelligent Network Assistant (NINA): A Multi-Lingual Chatbot API for Subscription and Order Services Using REST and GraphQL

This study presents Nice Intelligent Network Assistant (NINA), a multilingual chatbot designed to support subscription and order services via both REST and GraphQL APIs. Developed following the CRISP-DM methodology, NINA integrates natural language processing (NLP) techniques to recognize user intent, extract key entities, and generate accurate responses. The system supports multiple languages, including English, Xhosa, Shona and Ndebele, by translating user inputs into English using Google Translate before intent classification. A Bi-Directional Long Short-Term Memory (Bi-LSTM) model was trained on a Kaggle dataset to classify English-language intents with high precision. The model achieved 100% test accuracy with zero test loss, demonstrating its robustness and reliability. These results underscore NINA’s ability to deliver consistent and efficient support without human supervision. By combining deep language understanding with flexible API deployment, this work highlights the potential of multilingual AI assistants to enhance user experience and service delivery across diverse linguistic contexts.