Thailand’s higher education institution, Chulalongkorn University (Chula) has launched ChulaGENIE (Chula’s generative AI environment for nurturing intelligence and education) - a generative AI application to support research and learning for over 50,000 staff, faculty, and students.
It provides multilingual, multimodal capabilities, enabling users to customise AI agents for research, academic advising, and administrative support.
The application is built on Google Cloud's Vertex AI
ChulaGENIE will be initially available to Chula staff and faculty members in January 2025, with all students receiving access in March 2025.
It connects to Model Garden on Vertex AI, giving users access to a selection of AI foundation models designed for different tasks.
Model Garden on Vertex AI provides a single place to search, discover, and interact with a wide variety of models from Google and Google partners.
Currently, users can work with Google’s Gemini 1.5 Flash or Gemini 1.5 Pro.
Soon, users will be able to use enterprise versions of Anthropic’s Claude models and Meta’s Llama models.
With the Gemini models, users can understand complex topics or create content quickly in multiple languages, including Thai and English.
With these models’ multimodality and long context window, users can upload lengthy and complex documents, for instance, 1.4 million words with tables, charts, and images, including PDFs.
Gemini models can process text and visual elements to help users extract specific information or insights.
This is useful for researchers who need to summarise academic literature or analyse datasets.
Assistance for specific tasks
The university will introduce a new functionality within ChulaGENIE.
It will help members to create custom AI agents specialising in a particular area to assist with specific tasks.
Research assistants fine-tuned on niche topics, like carbon sequestration or urban traffic flow, help faculty and students identify connections and contradictions in research papers and propose new questions or hypotheses.
Academic advisors are trained on a knowledge base of academic and career information, student support resources, educational best practices, and ethical considerations.
This allows students to obtain tailored advice on course selection and career paths.
Administrative assistants who can comprehend and respond to queries in areas like admissions, enrollment, financial aid, facilities management, or IT support.
Chula focuses on responsible AI use and is using Vertex AI’s safety filters and its own governance framework to prevent harmful content in ChulaGENIE.
To improve response accuracy, it will also use Grounding with Google Search - a feature that enhances the accuracy and relevance of AI-generated responses by incorporating real-time information.
Google Cloud’s Thailand country director, Annop Siritikul, said, the Vertex AI platform helps Chula to operationalise responsible AI through features including grounding, model evaluation services, and tools for meeting stringent standards around data governance, privacy, and intellectual property protection.
Chula's rollout of ChulaGENIE along with a new Thai large language model for higher education and Google's AI skilling programs, will help in research and academia, Siritikul said.
The university uses Google Cloud’s encryption and access controls to protect user data and prevent leaks of sensitive information.
Open-source Thai LLM
The plan is to work with Google Cloud to create an open-source Thai large language model for higher education.
This will enable personalised learning and curriculum analysis based on the latest research and trends.
Additionally, Chula offers the Google AI Essentials course for free through Google's Samart Skills program, with over 800 participants earning certificates in generative AI tools.
The course includes expert videos, hands-on practice, and assessments to teach effective prompting and responsible AI use.
The university is also simplifying the curriculum for non-native English speakers.
This is to ensure everyone can use ChulaGENIE to enhance productivity, solve problems, and drive innovation.
