UBC LTIC has launched a new API service designed to support the development and use of AI-powered applications across the university. The service provides UBC community members with a centralized, institutionally supported way to connect their applications to large language models (LLMs), helping make AI integration more accessible and secure. Developed by the LTIC Incubator, the service responds to a need for reliable access to AI models, as faculty and staff express growing interest in incorporating AI into their work.
With AI-powered applications becoming more widespread, developers must navigate challenges related to privacy, security, and costs. These considerations are crucial when applications involve sensitive university data that must not be shared externally. LTIC’s API service addresses these issues by providing a central system through which UBC-developed applications can connect to language models hosted both at the university and third-party providers.
The API service has undergone UBC privacy assessments, which provides an additional level of assurance for faculty and staff looking to integrate AI capabilities into their projects. Applications involving sensitive data can benefit from access to models hosted within UBC, while other projects can connect to externally hosted models where appropriate. Developers can access language models through the service while maintaining the same API capabilities they would otherwise receive from external providers.
By reducing some of the technical, privacy, and financial barriers associated with AI development, the service aims to make it easier for members of the UBC community who are interested in incorporating AI technology into their work. LTIC’s API service supports real-world programs used by faculty and students, including several that are already in production, such as BIOCBOT.
LTIC’s API service is now available to UBC faculty and staff who have developed an AI-powered application or are currently building one. Those interested in using the service can visit the Large Language Model API Service page to learn more about its capabilities and how to request access.