Nanyang Technological University, Singapore (NTU Singapore) has partnered with AI provider CloudMile to enhance its student virtual assistant, Lyon Bot.
By integrating Google’s Gemini large language model (LLM), the upgraded AI chatbot promises faster, more accurate responses to student inquiries—especially on accommodation. This boosts engagement while streamlining administrative tasks and saving the university over 14 days of manual work each month.
“We are transitioning from an intent-based Natural Language Processing chatbot to a generative AI chatbot for student housing inquiries to enhance flexibility, scalability, and user experience.” said Ng Kee Haur, deputy director, Enterprise IT, Centre for IT Services at NTU.
“While generative AI can handle complex, unstructured queries, there is still a role for the intent-based chatbot in specific scenarios,” said Ng. “This hybrid approach allows us to balance efficiency with precision.”
NTU initially introduced Lyon Bot in 2020 to assist nearly 6,000 freshmen with enrollment, easing the administrative burden on university staff. The original system faced challenges adapting to evolving student needs and required laborious manual data tagging and updates.
With CloudMile’s expertise, NTU has now advanced Lyon Bot into a generative AI-powered system. The new version employs multi-layer processing — including query understanding, demand classification, and answer generation — to retrieve information in real-time from multiple centrally managed data sources. This has markedly improved conversational accuracy and response speed, significantly reducing error rates.
“With the new AI-powered chatbot, we are currently saving around 14.5 days’ worth of work every month,” said Alvin Ong, CIO of NTU. “This allows our staff to focus on tasks that require human insight and creativity.”
CloudMile provided customized consulting services and system integration tailored to the specific needs of the educational institution, helping NTU in building a stable and scalable AI conversational framework.
