Q: What datasets are you currently maintaining that aren't being adequately leveraged to generate value for your institution or organization? Why do you think those datasets are valuable and what problems do you think they might solve if they were used for AI modeling?
Hello all! Evan Claudeanos joining the conversation from Ellucian's incubation team. I haven't yet had a chance to introduce myself, but I have been watching this community, and it's great to see you all connecting over such pressing topics.
My team at Ellucian is actively developing products that combine new and existing datasets to power novel approaches to AI/ML. We're currently incubating a solution called Ellucian Journey, which leverages student and course data to build skill taxonomies, connect those data to recent workforce data, and help students identify opportunities to effectively advance their careers.
While I'm super excited about our work, I wanted to start a discussion focused particularly on identifying data opportunities. We know that great data is at the core of any strong AI/ML capability. (I sometimes worry that we've lost this thread in the era of pretrained general purpose LLMs like ChatGPT, Claude, and Llama, which can be dropped into workflows without much effort.) Ideal datasets tend to be large and should contain relevant, high-quality information with respect to the problem one is looking to solve.
I get the sense that we have many such datasets across institutions and organizations that are underutilized and underappreciated. Identifying such datasets has been one of my team's top priorities, which is why I wanted to put the question that I began with to all of you. I'd love to hear your ideas!
------------------------------
Evan Claudeanos
Director of Product Management
Ellucian
------------------------------
Message from the HEUG Marketplace:------------------------------
Find, Review, and Engage with Higher Education-focused solution providers, products, and services using the
HEUG Marketplace.
------------------------------