Carnegie Mellon University (CMU), a leading global research institution set out to solve a deceptively simple challenge: how to better capture and showcase the full breadth of its researchers’ outputs, without duplicating effort or hosting redundant work.

Their solution? A “Refer-atory.”

CMU’s research repository, KiltHub (a Figshare Instance), had, up until recently, focused on storing original content. But with the University’s impressive volume of externally published research, there was a growing need for a complementary approach; one that could create lightweight, metadata only records pointing to original sources, automatically, and at scale.

“ There is a lot of potential to leverage the data coming in from oaswitchboard to bolster other projects and add useful scholcomm data about our authors, publishers and publications..” Jonathan Kiritharan, Software and Applications Developer - CMU

Turning Metadata into Meaningful Visibility

Partnering with OA Switchboard allowed CMU to pilot a workflow that brings structured, publisher supplied metadata directly into KiltHub. By leveraging OA Switchboard’s API, the team at CMU created a script to query incoming messages and pull in key metadata that would otherwise be laborious to gather and manually input.

The OA Switchboard integration allowed CMU to test a novel repository model:

·       Automated “refer-out” records that point users to the official version of record

·       Create and test a script to format JSON data into KiltHub-friendly fields

·       Enabled SelfDOI functionality in Figshare to retain original DOIs

·       Tested workflow in KiltHub’s sandbox environment

While still in testing via their KiltHub sandbox, the model shows promise as a scalable, minimal effort way to boost research visibility and institutional-tracking without requiring their team to manually input full-text deposits.

Navigating the Integration: Lessons and Insights

The integration wasn’t without its hurdles. The CMU team had to tinker with metadata formats and navigated platform constraints, making sure the communication between OA Switchboard and Figshare flowed smoothly.

Despite these challenges the partnership between CMU and OA Switchboard has already been fruitful. The model now allows for communications that cut down on staff burden and ensures that much more of CMU’s institutional research is now hosted in their repository and may soon extend to even more metadata types.

Why this is Important

The CMU case reflects the wider shift in institutional repository strategy, from static storage to dynamic models that integrate the latest technology. Rather than duplicate or replace publisher records the “Refer-atory” model adds value by improving discoverability, and centralizing visibility across fragmented systems.

It is a practical response to a broader problem: too much research slipping through institutional nets as open access becomes more prevalent, and workflows become more labour intensive and disconnected.

OA Switchboard’s role in the pilot demonstrates its core strength – connecting systems and stakeholders through shared, structured infrastructure. It’s not about replacing what already works, it’s about connecting it.