No Migration Required: Decoupling Repository Innovation from Migration
The needs of digital collections continue to evolve, but the repositories that house them often cannot keep pace. That leaves institutions with a hard choice: migrate to a new platform to get new capabilities, or do without them. We’ve been exploring what it could mean to decouple innovation from repository migration, enabling institutions to improve how collections are prepared, enriched, and made accessible while continuing to use the platforms they already have.
The Migration Challenge
Migration can be a significant undertaking for a digital collections team, requiring budget, staff time, and technical capacity. For many institutions, those resources are limited or already stretched across competing priorities. At the same time, requirements around accessibility are changing, digital collections teams need more from the tools that support metadata and derivatives, and interest in AI and other emerging technologies is growing. When those capabilities do not reach existing repository platforms, a gap can emerge between what institutions have and what they increasingly need.
The Digital Collections Lifecycle Challenge
Much of the work that can help close this gap happens before content reaches the repository. Preparing collections can involve reviewing files, transforming metadata, creating derivatives, addressing accessibility gaps, and performing quality control. This work may span spreadsheets, shared drives, folders, and other disconnected tools, making collaboration, quality control, and preservation of institutional knowledge more difficult.
What If Innovation Didn't Require Migration?
Together, these challenges have led us to explore how we can support more of the digital collections lifecycle independently of the repository. We are developing platform-agnostic tools that work alongside existing repositories to transform files and metadata, create derivatives, improve accessibility, enrich description, and prepare collections for publication.
AI has a role to play, particularly in easing the burden of time-consuming tasks such as transcription, image description, and metadata enrichment. But AI-generated output still requires the expertise and judgment of the people who understand these collections. Our approach keeps practitioners in the loop, reviewing and approving AI-assisted work rather than handing decisions about their collections to AI.
As we develop these tools, conversations with the digital library community will help us understand where new approaches can make the greatest difference.
Join Our Session at the DLF Forum
We’ll be sharing an early look at this work at the virtual DLF Forum, October 14–15, 2026, which brings together practitioners from across the digital library community to share ideas, projects, and approaches to the challenges shaping their work.
CTO Noah Smith will present the lightning talk, “No Migration Required: Next-Generation Tools for Existing Repositories,” on Wednesday, October 14 at 2:30 p.m. ET.
If the challenges we’ve described sound familiar, we’d love to hear how your institution is approaching them and where better tooling could help. To learn more, contact Discovery Garden.
Agostino Ramelli’s book wheel, from Le diverse et artificiose machine (1588). Courtesy of the Library of Congress, via The Public Domain Review. Public domain.
