AI LABS: ENABLING STUDENTS AND ADMINISTRATORS OF INDIAN EDUCATIONAL INSTITUTES TO USE THE COMPUTE RESOURCES OF E2E NETWORKS

Abstract

This report, by author Vidushi Gupta, written under the mentorship of her guide Prof. Taruna Singh, is about AI Labs: Enabling Students and Administrators of Indian Educational Institutes to Use the Compute Resources of E2E Networks. This report documents the design process of this project which was undertaken as part of my final semester internship at the leading cloud and infrastructure provider- E2E Networks. The intent of AI Labs is to create a centralized platform for providing GPU compute resources to educational institutes of India to be, in turn, used by students, professors and university administrators country-wide. This platform is specifically for those domains that require compute for their practical learning, for example, domains like artificial intelligence, machine learning, data science, computational science, etc.

Since compute is crucial for learning in these booming educational domains, every institution must make arrangements for these, so as to enable their students to be able to learn better. During this project, it became clear that this is currently not the case. Compute infrastructure in the institutes is a limiting factor for several students. Even the ones that do have it have a very fragmented and unorganized provisioning for it. Consequently, students rely on personal systems or open-course tools (limited) or third party tools, which are sometimes unethical.

AI Labs was developed to address these concerns and make a unified platform that provides compute to institutes, enables them to use it effectively by providing suitable software for it, and addresses their concerns of student- professor collaboration for the same.

My role as an intern here was to design this platform end to end. This included understanding user needs through interviews and discussions with professors, students, and university admins, defining the information architecture, designing dashboards, creating wireframes, and iterating on prototypes based on feedback.