ORI-AI: A Practical Session on AI in Research Information Work | Between Possibility and Policy

Open Research Information
Workshop
10:00-15:00
SURF Utrecht, Room 2.04
Free
Register (to do)

Save the date! AI offers a lot for research information work (data enhancing, reporting, dashboarding, etc.), but privacy, security frameworks, and limited organizational agility often slow adoption down. This session isn't about debating whether AI is useful, but about arriving at shared working methods and a practical workflow for when and how to use it : covering everyday use of Claude and local LLMs, MCP servers as a way to connect AI to institutional tools, a demo of the SURF AI Hub, ORI DataLake Chatbot, and a hands-on discussion where we route real cases through a draft decision tree built around data sensitivity, scale, reproducibility, and auditability.

0 Praat mee

https://pixabay.com/photos/egyptian-goose-duck-goose-chick-5000132/

What to expect:

  • A walk-through of working with Claude, Claude Code, and local LLMs in practice,  including honest time estimates and a case where it didn't work
  • An introduction to MCP servers and how they connect AI to institutional tools like CRIS platforms and repository systems
  • A demo of the SURF AI Hub and ORI DataLake Chatbot as a middle path within institutional frameworks
  • A hands-on discussion block where we route real cases through a draft decision tree (data sensitivity, scale, reproducibility, auditability)
  • A concrete takeaway: one sentence, document, or pilot you can bring to your own institution board

 

Come with one task: please think of one concrete task you're currently unsure how to handle with AI, something you tried and abandoned, are considering, or a colleague asked about. We'll collect these on the day to stress-test the framework together.

 

Full agenda and pre-meeting details to follow closer to the date.

 

Organised by Nick Veenstra (RUG) and Maurice Vanderfeesten (SURF)

Dit artikel heeft 0 reacties