Welcome to AI4RA Tech Talk, A Shared Space for Building Technical Knowledge
By Barrie Robison
We are excited to launch our AI4RA Tech Talk blog series which is intended to complement AI Trail Blazer. Our goal is to facilitate the sharing of technical knowledge among our community of practice as we work together to develop practical and reliable tools for Research Administration.
To kick off the Tech Talk Series, we invite our readers to help us address a pressing question at the current stage of this project: How do we accelerate the dissemination of our AI tools?
Vandalizer is the framework we have been developing to help augment RA workflows with AI. When we mapped out the AI4RA project, we deliberately created a conservative timeline to release Vandalizer as open-source software. This gave us time to release well tested software with a high-quality codebase.
The recent presentations and demonstrations of our work have resulted in enthusiastic engagement and a high level of interest in trying out our tools. Our standards for Vandalizer remain high, but we also want to create pathways for people to engage with this tool as soon as possible.
We are interested in fielding ideas from our community of practice to help us create access points to Vandalizer, including development versions, prototypes, etc. Facilitating access during development helps us source feedback on potential features and bugs, informing better design as we work to make these tools reliable and practical in a variety of institutional conditions.
We are prioritizing this point of discussion among the AI4RA leadership team, our developers, and our university leaders to increase access to Vandalizer. Through these discussions, we have identified challenges related to data security, software support, training, and model access, among others. In our planning, we are working on a variety of solutions that allow us to reliably navigate these challenges while still accelerating access. Creating access is important, but we can’t introduce compromises on any of our core development principles, including Accuracy, Reproducibility, Flexibility, and Security.
The AI4RA community includes developers, analysts, data scientists, and other technical personnel who have a keen interest in this project. We want to connect with you and learn from your insights on AI development and access. How can we provide access to our tools without compromising our core principles? We invite your perspective in the comments section below!
And if you haven’t already, we encourage you to leave your information on our contact page for updates as our project develops.