Samvera Virtual Connect 2026

Samvera Community Wiki


Samvera Virtual Connect 2026

Watch Presentations on YouTube

Shared Community Notes Document

Virtual Connect is a free, informal opportunity to share work, ideas, and experiences with current and potential adopters of Samvera technologies. The event takes place entirely online via Zoom, and presentations will be recorded for asynchronous viewing so the broader community can learn from them.

Day 1 -  Tuesday, May 12, 11am - 2:00pm Eastern Daylight Time

Time

Title

Abstract

Presenters

Video/Slide Links

Time

Title

Abstract

Presenters

Video/Slide Links

11:00 - 11:30

Welcome; platform community updates; report from the Hyrax/Hyku Technical Coordinator

Updates from Product Owners for Avalon Media System (slides), Hyku, and Hyrax (slides), plus a presentation from the Hyrax/Hyku Technical Coordinator

@Jon Cameron , @Nicholas Mark Homenda , @Nic Don Stanton-Roark , @Rob Kaufman

https://youtu.be/KcbSdhPryuw

https://youtu.be/BfIquMwOyas

https://youtu.be/qOJiq-vvono

https://youtu.be/2VvMWpuGOZg

Hyrax Update slides

11:30 - 11:50

Developing VPATs for Hyrax and Hyku

In 2025, the Samvera Accessibility Working Group was charged with creating a Voluntary Product Accessibility Template (VPAT) for each community-maintained platform, prompted by many institutions’ policies requiring VPAT documents as part of software accessibility review processes. As members of this group, we facilitated work on VPATs for Hyrax and Hyku. Findings from working on these VPAT documents have informed prioritization of accessibility features and issues in Hyrax and Hyku, some of which have since been completed and incorporated.

This session will discuss our process, lessons learned en route to developing and releasing Hyrax and Hyku’s VPAT documents (including building upon prior accessibility audits), learning from the open source community, and ideas for keeping these documents current.

@Nicholas Mark Homenda , @Christine Peterson

Developing VPATs for Hyrax and Hyku slides

 

https://youtu.be/stYHGqWL-Ew

11:50 - 12:00

Break

 

12:00 - 12:20

Working Smarter AND Harder: How the Samvera community funded, designed, and built the Bulkrax CSV Guided Importer

CSV imports have been one of the most common and most painful parts of working with any bulk importer. Rather than solve it in isolation, Notch8 partnered with a coalition of Samvera institutions to fund, scope, and deliver the new CSV Guided Importer as a community initiative. This talk is a candid look at what happens when you bring a product-centric approach to open-source development: how the work was funded across institutions, how we defined outcome objectives, how librarians and repository managers shaped the design through structured discovery, how engineering and product management kept scope honest against the funding envelope, and what we learned along the way. Finally, we'll talk about how we intend to measure our success moving forward, and how that can inform our future investments. The goal is to show other Samvera institutions and partners what this playbook looks like in practice, whether the scope is a Bulkrax feature or something else entirely.

@Nick Steinwachs and Notch8’s Team Violet

https://youtu.be/waPtx3iOkyA

12:20 - 12:40

Flexible Metadata: Now, Next, and After That

Flexible Metadata has arrived — but where does it go from here?

In this session, we'll walk through the current state of the M3 profile system as implemented in Hyku and Hyrax. We'll share what near-term improvements are on the community's radar, and sketch out some bigger-picture thinking about what a mature, fully-realized Flexible Metadata feature could eventually look like. Come ready to weigh in — we want to hear from anyone using it, evaluating it, or with opinions about where it should go.

@Nic Don Stanton-Roark , @Rob Kaufman

https://youtu.be/JsQe3ifaS3A

12:40 - 12:50

Preserving Richness, Enabling Access: A Dual Repositories approach to migrating a complex Digital Collection

Change is the only constant, and nothing proves that more than migrating a complex digital collection into a new system. In this talk, we share the real-world challenges, and the unexpected discoveries that came with migrating the Slocum Puzzle Collection. Originally housed on IU’s Image Collections Online platform, the collection included more than 37,000 works, each with rich, sometimes messy metadata, plus a mix of single-image and multi-image records.
To honor this richness while supporting sustainable long-term access, we adopted a two-repository approach: one optimized for preserving the full descriptive depth of the archival metadata, and another that delivers a cleaner, more friendly view for users. This strategy allowed us to maintain the original collection context while meeting the structural requirements of our Hyrax-based access system. This meant making thoughtful choices about which fields could be combined, while preserving nuance without overwhelming the access system along with making decisions about redundant fields across the two repositories. Migration is always filled with unexpected challenges, and we ran into our share including inconsistent field usage, file matching, and obviously the data cleanup.
Ultimately, these challenges helped us build a user-friendly repository and a deeper appreciation for the collection itself.

@Sudha Anand & @Juliet Hardesty , Indiana University

Slides (pptx)

https://youtu.be/gmC7kbEtcjg

12:50 - 1:00

Break

 

 

 

1:00 - 1:30

Lightning Round: Repository showcases, group updates, and lightning talks

Repository Showcase: BTAA Hosted Streaming Media, Indiana University

Lightning talk: A Newer, Friendlier Bulkrax Importer (slides)

Updates from working and interest groups: AI Working Group

@Jon Cameron

@LaRita Robinson

@Nic Don Stanton-Roark

Bulk importer slides

https://youtu.be/LTQgCcWoqdQ

https://youtu.be/BioXUJB2s4U

 

1:30 - 2:00

Special Breakout Session: Who Decides? Building a Community Decision-Making Framework for Hyku

How does an open source community decide what gets built, who does the work, and how priorities get set — especially when contributors span institutions, consortia, and varying skillsets and levels of capacity? This session dives into the community decision-making framework and development workflow that the Hyku community has been developing as part of the IMLS-funded Sustaining Hyku grant project, in partnership with Invest in Open Infrastructure (IOI).

We'll briefly walk through the emerging framework, including a community leadership structure, a RACI diagram, and a draft Community Development Workflow — and then open the floor for discussion. We're looking for honest, practical feedback from across the Samvera community: Does this resonate with how the community works and how you see the community maturing? What's missing? Where are the sticking points?

Come ready to share your perspective. This session is targeted at anyone interested in how Hyku decisions are made, and/or who wants to contribute to the process. Also, anyone interested in how these ideas intersect with other Samvera platforms (e.g., Avalon, Hyrax) is welcome to attend!

Guiding questions for discussion:

  • How do you currently see yourself (or your institution) in the Hyku decision-making process?

  • If you are a Hyrax or Avalon user, how do you see your work being incorporated into the Hyku decision-making?

  • What would make it easier to contribute ideas or features to the community?

  • How should the community balance openness to contribution with the need for coordinated prioritization?

  • What would a well-defined maintenance and/or development fund mean for your institution's engagement?

Facilitated by @Amanda Hurford with others from the Sustaining Hyku Decisionmaking and Collaboration Group

Slides
Decision-making Framework documents
Brainstorm & Discussion Padlet
Sustaining Hyku IMLS Grant
Blogs on work w/ IOI for Decision-making:
PALNI and PALCI partner with IOI to build Hyku co-investment and decision-making model
Building toward sustainability: Partnering with the Hyku community on governance and funding

https://youtu.be/Z6VN-eIqnlY

 

 

Day 2 - Wednesday, May 13, 11am - 2:00pm Eastern Daylight Time

Time

Title

Abstract

Presenters

 

Time

Title

Abstract

Presenters

 

11:00 - 11:10

Repository showcase: West Virginia University

A showcase of the West Virginia & Regional History Center Digital History Collection

@Jessica McMillen & @Patrick Burden

 

11:10 - 11:20

Repository showcase: Oxford University

A showcase from Bodleian Libraries, University of Oxford

 

 

11:20 - 11:30

Sustaining Hyku: An IMLS Grant Update

This presentation provides a community update on Sustaining the Hyku Repository Platform: Addressing Hyku's Unique Community Coordination and Collaboration Challenges, a 2024–2026 project funded by the Institute of Museum and Library Services (IMLS) and led by PALNI and PALCI. Now in the second year of the grant, we'll share progress on key deliverables, highlight lessons learned, and preview what's ahead. Topics will include the partnership with Invest in Open Infrastructure (IOI) to research decision-making and financial sustainability models, advances in community development, including improvements to Bulkrax, and the work of our collaboration and finance subgroups to build a more resilient, community-driven future for Hyku. This is a high-level update for anyone curious about the project and how the Samvera Community can engage.

@Jill Morris , @Kirsten Leonard , @Amanda Hurford

 

11:30 - 11:40

We did it - Welcome to Fedora 7.x

Hyrax Fedora 6 Working Group Update

This presentation will discuss the path to Fedora 7.x, the first major version release since 2021. We will share what's new with this release and what it means for our users.

We will also discuss new ways to engage with the Fedora community and what we're planning for the future.

@Arran Griffith , Fedora

Slides

https://youtu.be/1Cp-qtYGMBc

11:40 - 11:50

Fedora as a Supplementary Repository in Hyrax

Can a repository be both fast and resilient? We propose using Postgres as the primary Hyrax metadata store with asynchronous synchronization to Fedora to get the best of both worlds.

@Daniel Pierce & @Randall Floyd , Indiana University

https://youtu.be/3dA2t6oCbWI

11:50 - 12:00

Break

 

 

 

12:00 - 12:20

Going Up - Keeping Your App Up to Date w/ Hyrax

This talk will go over tactics and suggestions on how to upgrade Hyrax. It will also introduce some new community resources and upcoming efforts around making upgrades easier and less risky.

@Rob Kaufman , Hyrax/Hyku Technical Coordinator

https://youtu.be/VOX7o9nJdvU

12:20 - 12:30

We Never Thought We’d Need to Do This: Locating and Preserving At-Risk Federal Publications in Hyrax

In February 2025, many datasets were removed from United States federal government websites. This alarming development led library staff at the University of North Carolina at Chapel Hill (UNC-CH) to worry that federally funded articles and research outputs might follow. To preserve UNC-CH authored research in service of UNC-CH's Open Access Policy, librarians rapidly developed and deployed the Federal Repository Harvesting Project. For this project, we identified UNC-CH authored research held in multiple federal repositories, harvested metadata and available PDFs, and uploaded the outputs to our Hyrax-based institutional repository.

In this presentation, I will describe the project, including our repository identification criteria, our methods of harvesting eligible content, and the ingest process. I will reflect on the lessons that our team learned as we deployed a project rapidly to address a continually changing issue. Finally, I will address the changes that we have made to our repository collection strategy to convert the project from a one-off to an ongoing process.

@Rebekah Kati , University of North Carolina at Chapel Hill Libraries

https://youtu.be/zrAajIY-GLk

12:30 - 12:50

Transforming RDF metadata descriptions with SPARQL CONSTRUCT

Academic libraries in the United States increasingly publish and consume collections data using the Resource Description Framework (RDF) model, as seen in expanding cataloger contributions to Wikidata, continuing adoption of BIBFRAME and other RDF schemas in production, and even the recording of Uniform Resource Identifiers (URIs) in MARC records.
The Oregon Digital Hyrax repository implements RDF by configuring a property URI for each metadata element and allowing metadata creators to record values from dozens of RDF data stores, including vocabularies published by the Library of Congress and the Getty Research Institute, and also enables collection-level RDF exports for administrators. These features represent years of effort as an early adopter of RDF for digital-collections metadata.
However, RDF exports are not suitable as-is for publication and integration on the web.
This talk presents a proof-of-concept utilizing the SPARQL query language, a core RDF standard, to transform exports into datasets ready for publication and reuse. This relies on SPARQL’s CONSTRUCT query form to transform an input graph into output conforming to an application profile or conceptual modelling scheme via a mapping in the query itself and utilizes the Python rdflib library to process the query and handle output.

@Benjamin Riesenberg , University of Oregon Libraries

https://youtu.be/-hujUFXvS7Q

12:50 - 1:00

Break

 

 

 

1:00 - 1:10

Interest Group Updates

Updates from the Documentation Interest Group and the Metadata Interest Group

@Morgan McKeehan , @Sarah Proctor

@Emma Beck , @Annamarie Klose

https://youtu.be/Tcxvm2ax2Ag

 

https://youtu.be/7WHNLRcyE6A

1:10 - 1:30

From Whack-a-Mole to Edge Protection Mitigating AI Scraping in Shared Repository Infrastructure (slides)

AI-driven scraping introduces a class of traffic that is low-volume but high-cost, targeting expensive endpoints in repository systems such as Hyku and Hyrax. In multi-tenant environments, this can lead to resource contention across shared services including Solr, PostgreSQL, and Fedora, resulting in degraded performance and cascading outages.

This presentation outlines an edge-based mitigation strategy using Cloudflare (free tier) to apply WAF rules, rate limiting, bot management, and caching before requests reach origin infrastructure. All configurations are managed via infrastructure as code, enabling consistent deployment across multiple environments with per-client customization.

We will detail implementation patterns, including selective allow-listing, behavioral rate limiting, and cache strategies, as well as operational lessons learned from real-world incidents. Attendees will gain practical insight into mitigating scraper-driven load, balancing access with protection, and managing edge controls at scale.

@April Rieger , Notch8

https://github.com/notch8/cloudflare-iac-example

 

https://youtu.be/H89E_KHrK9Q

1:30 - 2:00

Special Breakout Session: Hyrax Ingest - Could It Be Simpler?

The Valkerized ingestion process can be fairly complicated. If you add in iiif_print or the derivative rodeo work, it gets even more so. Though powerful, I'd like to go over the current set up and have a discussion around what we can do to simplify the process.

@Rob Kaufman , Hyrax/Hyku Technical Coordinator

https://youtu.be/tMOxSLxdFFI