The Business Education in Technology and Analytics (BETA) Hub at DePaul’s Driehaus College of Business has long been a place where faculty, staff, students, and alumni can experiment with integrating cutting-edge tools into their work — from eye-tracking software to a podcasting studio.  

Now, BETA Hub is debuting a new resource for Driehaus faculty and staff: an AI studio, complete with a high-performance computer capable of running large language models (LLMs) completely locally.    

The ability to run custom LLMs locally creates new avenues for faculty research, . As it opens up to Driehaus faculty and staff (and, under the guidance of at least one faculty mentor, students), we spoke with BETA Hub Co-Director Ignacio Luri about the new resource — and his insights into AI as a researcher who regularly works with large language models. 

Driehaus College of Business (DCOB): BETA Hub is launching a new AI studio! Tell us about what that looks like in practice.

Ignacio Luri (IL): BETA Hub's new AI studio, hosted at our space in DPC 7122, is BETA's way to support the college's AI@Driehus Initiative. Currently, BETA's AI studio includes one AI workstation: a high-performance computer capable of running Large Language Models (LLMs) completely locally. Other resources that the AI studio offers now or will offer soon are high-power cloud computing access, made-to-order agents and chatbots, and synthetic data generation.  

DCOB: How does the AI studio fit in alongside existing AI tools available to Driehaus faculty and staff? What kinds of projects is it well-suited to?  

IL: BoodleBox is great for teaching with AI; Copilot remains a good choice for everyday tasks. The AI studio supports more advanced use cases.  

Do you want to develop an AI app? Create an agentic framework that you control and own fully? Use generative AI to process unstructured data or automate tasks at scale? The AI studio can download the open-access LLM that fits your needs, customize it and run it locally. Nothing ever leaves this physical computer, or your DePaul account if you choose to log in, unless you want.

DCOB: You regularly work with LLMs in your research: for instance, to gain insight into the sentiments expressed in massive volumes of text. How does having a locally hosted LLM available in BETA Hub enhance your work?  

IL: For me, Generative AI's ability to process unstructured data is invaluable. Extracting themes or classifying large text databases, analyzing thousands of images, or processing and studying audio and video files: these are all things I’ve done, or want to do, as a researcher. In the past, that meant using less sophisticated methods, spending hours hand-coding data, or, most recently, using commercial APIs provided by AI companies (at a significant financial, and privacy, expense).  

Having this AI studio at Driehaus means that we now own the model whenever we want to do those tasks. No more token costs, limits or reading fine print for data safety red flags!  

DCOB: As someone who works with LLMs in your research, what's something you wish more people knew about AI?  

IL: This generation of AI comes with serious ethical issues. For me, the centralization of power in a few AI macro-corporations and the environmental impacts of AI are at the top of the list.  

Currently, my favorite innovation coming out in the AI space is the growth of relatively tiny, open-access Large Language Models. You would be surprised at what 4-6GB models can do on a decent laptop!  

The latest thinking model of the month is shiny — but we don't need it to write emails, summarize documents or brainstorm. Smaller local models and free, open-access AI are democratizing access, and they come at much less of a social cost.  

DCOB: How do you hope college faculty and staff will use the AI studio?  

IL: My mind goes to research because that’s where my AI expertise lies, but there are many ways to benefit from custom, fine-tuned LLMs that you can run locally. I'm sure there are people in the DePaul community experimenting with development or agentic productivity tasks far beyond my imagination!

The AI studio is now open to all faculty and staff, but we're also happy to support student projects that involve at least one faculty member. Reach out and set up a tour to better understand what the AI studio can do for you. Or if you know what you want, go ahead and reach us at betahub@depaul.edu letting us know what you want to achieve and your needs.

Ultimately, our goal is to enable AI-powered innovation and research at DePaul. Reach out if you want to help make that happen!