A finales de 2025, Invest in Open Infrastructure lanzó el Fondo de IOI para la Adopción en Redes, un nuevo tipo de inversión diseñado para acelerar la adopción global de infraestructura abierta para la investigación y el intercambio de datos. En lugar de financiar herramientas individuales o instituciones aisladas, el Fondo invierte en redes: consorcios de instituciones e investigadores con relaciones ya establecidas, una gobernanza activa de sus miembros y un compromiso demostrado con la ciencia abierta. La lógica es que fortalecer la infraestructura compartida que una red ya opera puede multiplicar el impacto en todas sus instituciones miembro al mismo tiempo.

El Fondo surgió del diálogo directo con las comunidades de investigación para identificar los proyectos de infraestructura abierta que más necesitan apoyo sostenido, y reúne contribuciones de un grupo de financiadores interesados en fortalecer el ecosistema global de investigación abierta. Cada red seleccionada recibe hasta 1,5 millones de dólares durante un período de dos a tres años, junto con apoyo continuo del equipo de IOI para la implementación en ámbitos como gobernanza, participación comunitaria y modelos de negocio. El objetivo de IOI con el Fondo no es solamente permitir que las organizaciones beneficiarias desarrollen proyectos capaces de generar un gran impacto en sus redes, sino también contribuir a la resiliencia de esas redes mucho después de finalizada la subvención.

LA Referencia fue seleccionada como una de las dos primeras beneficiarias del Fondo, entre más de 100 postulaciones provenientes de 22 países. En conjunto, las dos primeras organizaciones beneficiarias del Fondo están en condiciones de beneficiar a más de 1.300 instituciones de investigación en 36 países de América Latina y África. Este artículo ofrece una visión general de lo que LA Referencia ha realizado durante los primeros seis meses de esta inversión. En los próximos meses publicaremos también una actualización sobre la otra organización beneficiaria inaugural, UbuntuNet Alliance.

Introducción al proyecto de LA Referencia

LA Referencia es una red regional de cooperación que agrega y proporciona infraestructura de  acceso abierto a la producción académica y científica de repositorios nacionales de América Latina y España, basada en un modelo de gobernanza federada y en una infraestructura de propiedad pública y no comercial. Con el apoyo del Fondo de IOI, LA Referencia está desarrollando un proyecto para el período 2026–2028 destinado a ampliar el alcance regional y la profundidad técnica de esta infraestructura mediante cuatro grandes objetivos: incorporar más países y repositorios a la red, modernizar la forma en que los investigadores descubren contenidos en múltiples idiomas, construir un sistema soberano y verificable de identificadores persistentes y poner en marcha un repositorio regional compartido junto con un programa de capacitación para datos de investigación. El avance completo del proyecto está documentado en detalle en el sitio de informes del proyecto de LA Referencia, y el relato que sigue se basa directamente en el informe ejecutivo del equipo correspondiente a enero–julio de 2026.

El proyecto está organizado en tres componentes complementarios que conectan contenidos, tecnologías de descubrimiento, infraestructura de identificadores y capacidades humanas:

• Componente A — Expansión de la infraestructura de agregación de LA Referencia: ampliar la cobertura regional, mejorar y normalizar los metadatos e introducir búsqueda semántica multilingüe en la capa de descubrimiento de LA Referencia.
• Componente B — Identificadores ARK descentralizados y preservación de metadatos (dARK): desarrollar dARK 2.0, una infraestructura modular basada en blockchain e ipfs para asignar, publicar y resolver identificadores persistentes.
• Componente C — Repositorio regional de datos  datos para instituiones sin infraestructura local y soporte para Dataverse: poner en marcha un repositorio regional para datos de investigación basado en Dataverse, acompañado de un programa federado de capacitación y curaduría.

En el resto de este artículo profundizamos en los avances realizados en cada uno de estos componentes y compartimos productos y resultados del trabajo de LA Referencia que ya pueden ser explorados.

Componente A: Expansión de la infraestructura de agregación de LA Referencia

El descubrimiento depende de contar con buenos metadatos para las colecciones, por lo que el Componente A combina la ampliación de la cobertura regional de LA Referencia con el trabajo sobre calidad de metadatos y una nueva capacidad de búsqueda multilingüe que permite descubrir contenidos independientemente del idioma utilizado en la búsqueda o del idioma del contenido.

Avances hasta la fecha

Durante el período del informe, LA Referencia implementó una metodología regional para identificar instituciones y repositorios candidatos a ser cosechados con el objetivo de ampliar su colección. Este trabajo incluyó la verificación de servicios OAI-PMH, la revisión de metadatos, la actualización de la Interoperability and Registry Database (IRD) de COAR y la verificación cruzada de información institucional con ROR. El trabajo de curaduría se completó para los países miembros Ecuador, Uruguay y Costa Rica, y para los países no miembros Colombia, Paraguay, Venezuela, Bolivia y Cuba; los registros de otros países se encuentran en proceso.

En total, el equipo actualizó los registros de 362 repositorios: 187 fueron editados, 77 incorporados como nuevos y 98 archivados. Venezuela fue el país con mayores avances, con dos repositorios cosechados exitosamente como experiencia piloto para la incorporación de contenidos provenientes de países que todavía no operan como nodos completos de LA Referencia.

En relación con esta expansión geográfica, LA Referencia está desarrollando una solución para un problema común identificado en su comunidad: muchos investigadores realizan búsquedas en un idioma mientras que el trabajo que necesitan está descrito en otro. Un registro de un repositorio en español podría no aparecer nunca para una persona que realiza una búsqueda en inglés, o viceversa, porque la búsqueda tradicional por palabras clave solo puede hacer coincidir las palabras exactas presentes en la página. El Componente A busca cerrar esa brecha mediante búsqueda semántica multilingüe, que encuentra resultados haciendo coincidir significados en lugar de palabras exactas. Para determinar la mejor manera de hacerlo, el equipo realizó una evaluación rigurosa y reproducible, comparando aproximadamente treinta modelos diferentes de IA en dos colecciones de prueba de entre 30.000 y 50.000 documentos, con miles de búsquedas de prueba en hasta 10 idiomas. Este trabajo produjo resultados prometedores: en el piloto, la búsqueda semántica y la búsqueda combinada recuperaron aproximadamente el doble de resultados relevantes entre los primeros 10 resultados —pasando de un promedio de 4,5 a 9,1— en comparación con la búsqueda únicamente por palabras clave. Las mayores mejoras se observaron en idiomas que suelen presentar mayores dificultades para la búsqueda por palabras clave, como chino, japonés y árabe. El equipo evaluó distintos modelos y seleccionó uno que ofrecía un equilibrio entre resultados de alta calidad, eficiencia y velocidad para operar sobre millones de registros.

Esta evaluación ya fue incorporada a un sistema beta funcional dentro de la plataforma de búsqueda de LA Referencia. Cuando una persona realiza una búsqueda, puede elegir entre la coincidencia tradicional por palabras clave, una coincidencia “basada en significado” que funciona entre distintos idiomas o un modo combinado que integra ambas aproximaciones. La reindexación se vuelve considerablemente más rápida cuando se reutilizan embeddings generados previamente. Reconstruir el índice completo para un conjunto de prueba de aproximadamente 28.000 registros requirió alrededor de 5 horas la primera vez, pero solamente unos 10 minutos cuando los datos basados en significado ya estaban almacenados, lo que hace posible realizar actualizaciones regulares a medida que la colección continúa creciendo. Paralelamente, el equipo incorporó un WAF abierto que filtra bots automatizados y scrapers, junto con una forma de aplicar esa misma protección simultáneamente en múltiples sitios de LA Referencia.

También es posible profundizar en los detalles técnicos consultando el informe completo del Componente A de LA Referencia.

Próximos pasos

LA Referencia continúa revisando e incorporando repositorios de otros países miembros y de nuevos países fuera de su red. El equipo planea ampliar la evaluación de la búsqueda semántica multilingüe a toda la colección regional y consolidar la versión beta de búsqueda semántica antes de avanzar hacia un despliegue más amplio.

Componente B: Identificadores ARK descentralizados (dARK) y preservación de metadatos

El Componente B comprende el desarrollo de dARK 2.0, la próxima generación de la infraestructura de LA Referencia para asignar, publicar y resolver identificadores persistentes ARK. Este componente contempla una infraestructura regional, soberana y sostenible, construida específicamente para la región a la que sirve LA Referencia.

Avances hasta la fecha

Los identificadores persistentes son los que permiten que una producción de investigación continúe siendo localizable y citable a largo plazo mediante un enlace permanente que sigue conduciendo al artículo, conjunto de datos o informe correcto años después, incluso si el archivo cambia de ubicación o si cambia el sistema que lo aloja. Actualmente, gran parte de esta infraestructura es gestionada por un pequeño grupo de proveedores externos. dARK 2.0 representa la visión de LA Referencia de una infraestructura que la propia región pueda operar, gobernar y considerar confiable, sin depender de una única organización como centro de control.

Para lograrlo, el equipo dividió dARK 2.0 en componentes separados que pueden ser verificados, actualizados o escalados de manera independiente. Un registro compartido y resistente a manipulaciones —un tipo de blockchain— mantiene el registro autoritativo de qué identificadores existen y hacia dónde apuntan, mientras que los metadatos descriptivos de cada registro se almacenan por separado, permitiendo que cualquiera pueda verificar que no han sido alterados. Esta separación mantiene la capa de “prueba” simple y confiable, mientras que los datos descriptivos pueden continuar creciendo sin necesidad de modificar esa capa de “prueba”. El equipo también construyó todo el flujo de procesamiento que requiere el sistema: reservar nuevos identificadores, preparar y almacenar sus metadatos, publicar la referencia en el registro y permitir que cualquier persona consulte un identificador y lo resuelva hacia el registro correcto.

El equipo sometió este flujo de trabajo a pruebas de estrés con millones de registros y ejecutó cientos de verificaciones automatizadas para confirmar su robustez, además de construir herramientas administrativas para que el personal pueda monitorear identificadores, actividad y consultas. El flujo de dARK ya está listo para ser desplegado y conectado con el sistema regional de cosecha de LA Referencia, por lo que el proceso de cosecha se está integrando con el flujo de dARK para permitir la asignación de identificadores a medida que los registros son procesados, incluso en colecciones grandes y en constante cambio.

Profundizar en los detalles en el informe completo del Componente B.

Próximos pasos

El siguiente paso consiste en distribuir la infraestructura dARK entre distintas instituciones en lugar de ejecutarla desde un único lugar. Se está preparando una instalación equivalente en RNP, en Brasil, y la institución socia IBICT está preparando la migración de sus identificadores desde la versión original 1.0 de dARK. Con múltiples instituciones utilizando el sistema, LA Referencia planea validar su funcionamiento entre distintos sitios y formalizar el modelo de seguridad y gobernanza de esta infraestructura regional federada.

Componente C: Repositorio regional para datos sin infraestructura institucional y soporte para Dataverse

El Componente C combina un nuevo repositorio regional basado en Dataverse con un programa federado de capacitación y curaduría, cuyo objetivo es ofrecer un espacio para aquellos datos de investigación que carecen de infraestructura institucional, al mismo tiempo que se difunden las capacidades necesarias para operar servicios de datos de manera sostenible en toda la región.

Avances hasta la fecha

El equipo comenzó con un diagnóstico regional que identificó aproximadamente 96 repositorios de datos registrados y alrededor de 30 instalaciones de Dataverse en América Latina, concentradas principalmente en Brasil, Argentina, México y Colombia. El diagnóstico se basó en cuestionarios y entrevistas que abordaron gobernanza, infraestructura, curaduría, preservación, soporte e interoperabilidad. Este análisis confirmó que la región ya cuenta con una experiencia y conocimientos significativos, lo que dio forma a una estrategia de colaboración entre pares que concibe al repositorio como un servicio sociotécnico: software e infraestructura combinados con políticas, equipos multidisciplinarios, apoyo a investigadores y capacitación continua.

El principal resultado técnico es una versión de prueba funcional y en línea del repositorio —un “Dataverse Alpha”— que funciona sobre la propia infraestructura de LA Referencia. El repositorio está disponible en inglés, español y portugués, incluye campos para indicar el país de origen de un conjunto de datos y con cuáles de los 17 Objetivos de Desarrollo Sostenible de las Naciones Unidas se relaciona, y utiliza licencias abiertas (CC0 y CC BY) para que los datos puedan ser reutilizados libremente. Ya se han cargado conjuntos de datos de ejemplo para que el equipo pueda probar la carga, descripción, búsqueda y acceso a archivos antes de que comiencen a recibirse depósitos institucionales reales.

Junto con el desarrollo técnico, el equipo está elaborando un MOOC asincrónico de 16 horas, Introducción a la Gestión y Curaduría de Datos en Dataverse, que abarca fundamentos y publicación, metadatos y documentación, verificación técnica y formatos, y ética, licenciamiento y aprobación. Este es uno de los tres productos de capacitación planificados; los otros incluyen capacitación práctica en curaduría y apoyo directo a las instituciones piloto. El curso será construido e instalado en un entorno Moodle de producción, con materiales en español y portugués.

Invitamos a los lectores a explorar todos los detalles técnicos en el informe completo del Componente C.

Próximos pasos

El equipo finalizará los contenidos del curso, validará de extremo a extremo los flujos de trabajo del repositorio, seleccionará las primeras instituciones piloto y transformará las lecciones aprendidas durante la fase Alpha en un modelo estable para el servicio regional.

El proyecto de LA Referencia: apoyando los objetivos del Fondo de IOI para la Adopción en Redes

A seis meses de iniciado el proyecto, LA Referencia ha pasado del diseño y el trabajo de diagnóstico a sistemas funcionales: una experiencia de descubrimiento multilingüe con mejoras medibles, una arquitectura funcional para identificadores persistentes soberanos y verificables, y un repositorio regional de datos en funcionamiento acompañado de un proceso de capacitación diseñado para perdurar más allá de una única subvención. Nada de esto fue construido para una sola institución; cada mejora contribuye a una infraestructura compartida y federada de la que ya dependen los repositorios miembros de LA Referencia en toda América Latina.

Esta es la premisa del Fondo de IOI para la Adopción en Redes: que la forma más rápida y duradera de hacer crecer la adopción de infraestructura abierta es invertir en redes que ya cuentan con la gobernanza, la confianza y el alcance necesarios para poner nuevas capacidades al servicio de muchas instituciones simultáneamente. Los primeros seis meses de LA Referencia muestran esta tesis en acción. La base técnica que están construyendo permitirá avanzar hacia la próxima fase del proyecto: expandirse hacia nuevos países e idiomas, validar dARK entre distintas instituciones e incorporar las primeras instituciones piloto al Dataverse regional.

Leé más sobre el proyecto en el sitio de informes del proyecto de LA Referencia, incluido el informe ejecutivo completoen el que se basa este artículo, o conocé más sobre el trabajo de IOI con LA Referencia y el Fondo de IOI para la Adopción en Redes.

In late 2025, Invest in Open Infrastructure launched the IOI Fund for Network Adoption, a new kind of investment designed to accelerate the global adoption of open infrastructure for research and data sharing. Rather than funding individual tools or single institutions, the Fund invests in networks: consortia of institutions and researchers with established relationships, active member governance, and a demonstrated commitment to open science. The logic is that strengthening the shared infrastructure that a network already runs can multiply impact across every member institution at once.

The Fund grew out of direct engagement with research communities to identify the open infrastructure projects most in need of sustained support, and it pools contributions from a group of funders all interested in strengthening the global open research ecosystem. Each selected network receives up to $1.5 million over two to three years, paired with ongoing implementation support from IOI staff on governance, community engagement, and business modeling. IOI’s goal with the Fund is not only to enable grantees to do projects that can have a big impact on their networks, but also to support the resilience of these networks long after the grant ends.

LA Referencia was named one of the Fund's two inaugural grantees, selected from more than 100 applications spanning 22 countries. Together, the Fund's first two grantees are positioned to benefit more than 1,300 research institutions across 36 countries in Latin America and Africa. This post provides an overview of what LA Referencia has done with the first six months of that investment. Look for an update post about our other inaugural grantee, UbuntuNet Alliance, in the coming months. 

An introduction to LA Referencia’s project

LA Referencia is a regional cooperation network that aggregates and provides open access to the scholarly and scientific output of national repositories across Latin America and Spain, built on a model of federated governance and publicly owned, non-commercial infrastructure. With IOI Fund support, LA Referencia is running a 2026–2028 project to expand that infrastructure's regional reach and technical depth with four major goals: bringing more countries and repositories into the network, modernizing how researchers discover content in multiple languages, building a sovereign and verifiable system of persistent identifiers, and standing up a shared regional repository and training program for research data. The project's full progress is documented in detail on LA Referencia's project reporting site, and the account below draws directly on the team's own executive report for January–July 2026. 

The project is organized into three complementary components that connect content, discovery technology, identifier infrastructure, and human capacity:

• Component A — Expanding LA Referencia’s Aggregation Infrastructure: growing regional coverage, improving and normalizing metadata, and introducing multilingual semantic search to LA Referencia's discovery layer.
• Component B — Decentralized ARK Identifiers and Metadata Preservation (dARK): building dARK 2.0, a modular, blockchain-backed infrastructure for assigning, publishing, and resolving persistent identifiers.
• Component C — Regional Orphan Data Repository & Dataverse Support: standing up a Dataverse-based regional repository for research data alongside a federated training and curation program.

In the rest of this post, we dive deeper into the progress made on each of these components and share explorable outputs of LA Referencia’s work to date. 

Component A: Expanding LA Referencia’s Aggregation Infrastructure

Discovery depends on good metadata for the collections, so Component A pairs the expansion of LA Referencia's regional coverage with metadata quality work and a new multilingual search capability that enables discovery regardless of the language of the search or the content.

Progress to date

Over the reporting period, LA Referencia enacted a regional methodology for identifying candidate institutions and repositories for harvesting to bolster their collection. This work involved verifying OAI-PMH services, reviewing metadata, updating COAR's Interoperability and Registry Database (IRD), and cross-checking institutional information against ROR. Curation work was completed for member countries Ecuador, Uruguay, and Costa Rica, and non-member countries Colombia, Paraguay, Venezuela, Bolivia, and Cuba; records for another countries are in progress. 

In total, the team updated records for 362 repositories: 187 edited, 77 newly added, and 98 archived. Venezuela was the country with the greatest progress, with two repositories successfully harvested as a pilot for incorporating content from countries that do not yet operate as full nodes in LA Referencia. 

Related to this geographic expansion, LA Referencia is building a solution for a common problem they encountered in their community: many researchers search in one language while the work they need is described in another. A record from a Spanish-language repository might never surface for someone searching in English, or vice versa, because traditional keyword search can only match the exact words on the page. Component A set out to close that gap with multilingual semantic search, which finds results by matching meaning instead of exact wording. To figure out the best way to do this, the team ran a rigorous, reproducible evaluation, comparing roughly thirty different AI models across two test collections of 30,000 to 50,000 documents, with thousands of test searches in as many as 10 languages. This work resulted in promising outcomes: in the pilot, semantic and blended search surfaced roughly twice as many relevant results in the top 10 (up from an average of 4.5 to 9.1) compared with keyword search alone, with the biggest gains for languages that keyword matching tends to handle poorly, like Chinese, Japanese, and Arabic. The team evaluated different models and chose one that balanced high-quality results with efficiency and speed to run on millions of records. 

That evaluation has now been built into a beta working system on LA Referencia's search platform. When someone runs a search, they can choose the traditional keyword match, a "meaning-based" match that works across languages, or a blended mode that combines both. Reindexing becomes substantially faster when previously generated embeddings are reused. Rebuilding the full index for a test set of about 28,000 records took roughly 5 hours the first time, but only about 10 minutes once the meaning-based data was already stored, making regular updates practical as the collection keeps growing. Alongside this, the team added a open WAF that screens out automated bots and scrapers, plus a way to apply that same protection across multiple LA Referencia sites at once.

You can also dive into the technical details by reading LA Referencia’s full Component A report.

What's next

LA Referencia continues to review and add repositories from additional member countries and new countries beyond their network. The team plans to expand its multilingual semantic search evaluation across the full regional collection and consolidate the semantic search beta ahead of broader rollout.

Component B: Decentralized ARK Identifiers (dARK)  and Metadata Preservation

Component B involves building dARK 2.0, the next generation of LA Referencia's infrastructure for assigning, publishing, and resolving persistent ARK identifiers.  This component accounts for regional, sovereign, and sustainable infrastructure built specifically for the region served by LA Referencia.

Progress to date

Persistent identifiers are what keep a piece of research findable and citable for the long term by using a permanent link that still leads to the right paper, dataset, or report years later, even if the file moves or the system hosting it changes. Today, most of that infrastructure is run by a handful of external providers. dARK 2.0 is LA Referencia's vision for an infrastructure the region can run, govern, and trust for itself without relying on one single organization as a locus of control.

To do that, the team split dARK 2.0 into separate components that can each be checked, updated, or scaled independently. A shared, tamper-evident ledger (a type of blockchain) maintains the authoritative record of which identifiers exist and what they point to, while the descriptive metadata for each record is stored separately, allowing anyone to verify it hasn't been altered. That split keeps the "proof" layer simple and trustworthy, while the descriptive data can keep growing without needing to touch the "proof" layer. The team also built the full pipeline this system requires: reserving new identifiers, preparing and storing their metadata, publishing the reference to the ledger, and allowing anyone to look up an identifier and resolve it to the correct record.

The team stress-tested this workflow against millions of records and ran hundreds of automated checks to confirm it holds up, and built administrative tools so staff can monitor identifiers, activity, and lookups. The dARK pipeline is now ready to be deployed and connected to LA Referencia's regional harvesting system, so the harvesting workflow is being integrated with the dARK pipeline to support identifier assignment as records are processed, even across large, constantly changing collections. 

Dive into the details in the full Component B report.

What’s next

The next step is spreading the dARK infrastructure across institutions rather than running it from one place. A matching setup is being prepared at RNP in Brazil, and partner institution IBICT is preparing to migrate its identifiers over from the original 1.0 version of dARK. With multiple institutions using the system, LA Referencia plans to validate cross-site operation and formalize the security and governance model for this federated regional infrastructure.

Component C: Regional Orphan Data Repository & Dataverse Support

Component C pairs a new Dataverse-based regional repository with a federated training and curation program, aimed at giving research data lacking institutional infrastructure a home while spreading the skills needed to run data services sustainably across the region.

Progress to date

The team began with a regional diagnostic that identified roughly 96 registered data repositories and around 30 Dataverse installations across Latin America, concentrated in Brazil, Argentina, Mexico, and Colombia, drawing on questionnaires and interviews covering governance, infrastructure, curation, preservation, support, and interoperability. That diagnostic confirmed the region already has significant experience and expertise, which has shaped a peer-collaboration strategy treating the repository as a socio-technical service: software and infrastructure paired with policy, multidisciplinary teams, researcher support, and ongoing training.

The headline technical result is a live, working test version of the repository (a "Dataverse Alpha") running on LA Referencia's own infrastructure. The repository works in English, Spanish, and Portuguese, includes fields to tag a dataset's country of origin and which of the UN's 17 Sustainable Development Goals it relates to, and uses open licenses (CC0 and CC BY) so the data can be freely reused. Sample datasets are already loaded so the team can test uploading, describing, searching, and accessing files before real institutional deposits start coming in.

Alongside the technical build, the team is developing a 16-hour asynchronous MOOC, Introduction to Data Management and Curation in Dataverse, covering foundations and publication, metadata and documentation, technical verification and formats, and ethics, licensing, and approval. This is one of three planned training products; the others include hands-on curation training and direct support for pilot institutions. The course will be built and installed on a production Moodle environment with materials in Spanish and Portuguese.

We encourage readers to explore the full technical details in the full Component C report.

What’s next

The team will finish course content, validate repository workflows end-to-end, select the first pilot institutions, and turn the lessons from the Alpha into a stable model for the regional service.

LA Referencia’s Project: Supporting the Goals of the IOI Fund for Network Adoption

Six months in, LA Referencia has moved from design and diagnostic work to functioning systems: a measurably better multilingual discovery experience, a working architecture for verifiable, sovereign persistent identifiers, and a live regional data repository paired with a training pipeline built to outlast any single grant. None of this was built for one institution; every improvement contributes to shared, federated infrastructure that LA Referencia's member repositories across Latin America already rely on. 

That is the premise of the IOI Fund for Network Adoption: that the fastest, most durable way to grow open infrastructure adoption is to invest in networks that already have the governance, trust, and reach to put new capacity to work across many institutions at once. LA Referencia's first six months show that thesis in motion. The technical foundation that they are building will enable the project's next phase, expanding to new countries and languages, validating dARK across institutions, and bringing its first pilot institutions onto the regional Dataverse.

Read more about the project on LA Referencia's project reporting site, including the full executive report this post is based on, or learn more about IOI's work with LA Referencia and the IOI Fund for Network Adoption. 

Duration: April 2026 to December 2026
Team: Emmy Tsang, Emma Green, Sarah Lippincott
Funder: TDCC-NES
Skillset: Strategic Planning, Convening / Facilitation, Fiscal Assessment & Planning

Overview

TDCC-NES is the Thematic Digital Competence Centre for the Natural & Engineering Sciences in the Netherlands, a network organisation that helps Dutch researchers and digital support experts explore digital competency challenges and find solutions through connection and collaboration, reducing fragmentation and duplication across the domain. Funded by NWO as part of its investment in digital research infrastructure, TDCC-NES engaged IOI to develop a credible long-term sustainability model ahead of the end of the NWO funding window.

What we will do

Through a phased engagement, IOI is building a shared, evidence-based understanding of TDCC-NES's value proposition, stakeholder landscape, and sustainability options, through stakeholder interviews, document review, and a discovery findings presentation to the governing board. In late 2026, we will convene key stakeholders in a sustainability workshop to co-develop the foundations of a co-investment model and build co-ownership of the path to 2030.

Outcomes

Phase 1 delivered a discovery findings report and board presentation that established a shared evidence base for the board's sustainability conversation. Further outcomes will be added as the engagement progresses.

Facing a funding transition? Let's design your next model together.

Many publicly funded infrastructure and network organisations face the end of an initial grant or funding window. IOI helps organisations like yours understand your value to stakeholders and design sustainable co-investment models. Book a free consult.

Data resources that researchers rely on are critical to advancing science and discovery. That research often starts with data that research funders, such as Wellcome, did not create and does not maintain, including biodata repositories, population cohort databases, and environmental and climate data resources, built and sustained by organizations that often operate out of public view. At a moment when the research community is paying closer attention than ever to the resilience of the data infrastructure it depends on, we are supporting Wellcome to gain a clearer, deeper picture of that landscape. In turn, this will enable collective conversations within the research community about how we sustain these resources.

IOI is leading that effort, drawing on our track record of working alongside infrastructure providers, not just evaluating them from the outside. We'll be engaging directly with the teams running these resources to understand their governance, funding models, technical dependencies, open policies, sustainability outlooks and the broader ecosystem they operate within, the kind of insight that only comes from sustained, trust based engagement.

The foundation for this work is Infra Finder, IOI's open discovery platform for research infrastructure. We'll build verified, publicly accessible profiles for each resource in scope, using our established research and diligence process.This is also a significant expansion for Infra Finder, which has grown to include nearly 150 profiles of open infrastructure since its 2024 launch. Infra Finder is more than a directory. It's the foundation of our annual State of Open Infrastructure report, which tracks how the field is evolving year over year. Adding this cohort brings these data resources into that same yearly monitoring and analysis, giving Wellcome and the broader community longitudinal visibility into how they change and how they're funded over time.

We're looking forward to the work ahead. Stay tuned for more in the coming months!

Funders have responded to the crisis, with the Robert Wood Johnson Foundation, Hewlett Foundation, David and Lucile Packard Foundation, John D. and Catherine T. MacArthur Foundation, Alfred P. Sloan Foundation, Mellon Foundation, and others directing significant resources toward data rescue and resilience efforts. Some funders are working together in formal coordination groups (e.g., Funders for the Future of Public Data (3FPD) and the Portfolio to Protect Science) in hopes of moving resources in ways that reduce duplication, “noise,” while increasing investment alignment.

But a close look at where that money is flowing reveals concerning patterns that could subvert or undercut this work, limiting its ability to outlast the crisis that prompted it. Concretely, this includes dangerously low investments made in open infrastructures – including the technical platforms, standards, and software required to produce, share, discover, access, and preserve knowledge.

Crisis reveals what was always true

Funding is moving in familiar, system-driven ways that we have witnessed many times before. As my colleagues and I have noted elsewhere (e.g., Red Queen’s Race, Emperor’s New Clothes), stakeholders in the knowledge and information ecosystem have for years been rewarded for ways of working that provide short-term gains but that tend to fail to add up to long-term success.

Incentive structures, including funding and career advancement, encourage stakeholders to pursue innovation over maintenance, to build competing siloes rather than grow interoperability and coordination, and to misinterpret bare survival of open infrastructure communities and services as though that survival somehow signifies success.

The resulting fragility of open infrastructure components has been persistent, and the current political crisis of 2025-2026 hasn’t created precarity so much as it has exposed it. Data is disappearing because the infrastructure in which it is nested was already brittle. This moment is an opportunity, not just to rescue data, but to correct a long-standing dependency on temporary and inadequate funding for infrastructure-level work.

Tracking investment flows

IOI has been mapping the emerging data resilience ecosystem, tracking more than 85 “data rescue” and “data resilience” projects, initiatives, forums, and funding efforts that have launched or significantly evolved since 2025. The landscape clusters into recognizable functional layers: active rescue and collection efforts; tools, repositories, and discovery systems; monitoring and tracking; coordination and strategy; and advocacy and fundraising. Across these clusters, coordination-style work — convenings, roadmaps, governance documentation, advocacy campaigns — dominates the funded project list. Rescue and collection efforts, including some remarkable grassroots mobilizations, are also well represented in the list.

What is almost entirely absent in the projects and initiatives we are tracking is investment in the technical infrastructure layer:  the tools, pipelines, standards, systems, and people that make any of the other work durable. Of the 87 projects and initiatives we are tracking, 34 involve coordination activities, 28 involve advocacy, and 33 involve data collection, while only 16 are actively building and maintaining technical tools, and more than half of those are building for their own project's needs rather than building shared, reusable infrastructure. 

Those who are conducting “data rescue” or aiming for “data resilience,” depend on tools, pipelines, and standards to do their work, but those elements often are not being resourced through these funding streams. Rescued data has to be collected, and then it has to go somewhere; it has to be validated, described, made findable, and kept accessible. The systems that do that work, from web archiving tools to identifier infrastructure, from metadata standards to format normalization, and from discovery layers to storage and preservation activities, are either assumed to already exist and be adequately supported, or they are simply not visible in the investment picture. 

The danger is hidden: much of the content that is being rescued is flowing onto known and trusted infrastructures, e.g., Internet Archive, ICPSR and Data LUMOS, and Texas Advanced Computing Center (TACC). The longevity of these environments is unknown and they are not being funded to take on a permanent or future-facing role for this content. What does that mean for the maintenance and reliability of these resources over time?

The infrastructure layer is chronically underfunded relative to the work it enables, with costs hidden through volunteerism and institutional subsidy. The data resilience funding landscape is replicating that failure under crisis conditions.

Learning from previous experiences

This is, of course, far from the first time activist communities have mobilized to rescue endangered digital collections at scale. For example, the 2016 data rescue movement, responding to the first round of federal data threats, activated DataRefuge, Preserving Electronic Government Information (PEGI), and the Environmental Data and Governance Initiative (EDGI), among other efforts. Libraries, cultural heritage professionals, and researchers have also mobilized around a myriad of events over time, from the Estonian cyberattacks of 2007 to the conflict zones in Sudan, Syria, Ukraine, and Afghanistan; and from responses to the British Library ransomware attack of 2023 to rescue efforts in the wake of wildfires, floods, and other major natural disasters. Each of these efforts generated hard-won knowledge about what works — distributed custody models, community triage processes, the limits of dark archiving sans deliberate attention to access, and the differences between rescue and preservation. 

That knowledge is not being systematically drawn on in many of today’s projects, and the current US focus is producing efforts that are, in some cases, re-learning lessons that have already been learned elsewhere. We need to actively encourage and fund work that connects today's actors to this history. 

For example, do we know what parts of the 2016-17 data rescue tooling and environment still exist, and what parts do not? Do we know what characteristics differentiate still-active groups like EDGI, ICPSR, and Internet Archive from other groups and toolsets that have sunset? Knowing more about the past can help us to invest more wisely and sunset more adeptly, establishing stronger scaffolding for the field. That requires directing some portion of the funding towards practical analysis of these past efforts. 

Duplication of backbone (human) infrastructure

Our analysis shows a third gap that is perhaps the most costly today: instead of reinforcing the successful networks that already exist and distilling models from those that we can activate in other fields and disciplines, much of the funding available today is flowing to external contractors and consultants to build landscape analyses in order to better understand where the connectors and touchpoints between current efforts are. 

Unintentionally, these external groups are both duplicating and undermining long-standing, often mature or maturing networks on the ground. While the contractors and consultants entering this space for the first time gain remarkable resourcing, existing network players, including strong hubs of activation, struggle for survival, with staff living from contract to contract, and no organizational runway available to them to enable the strategic planning or expansion work they may be uniquely suited to do. In other words, rather than directing funds to strong cornerstone players (e.g., EDGI, PEDP) in support of the (volunteer-driven) bridge building they already have underway, resources are going to external and/or new groups.

This problem is structural, not intentional, but it still deletes resources from the grassroots actors and specialists in the field in at least three ways. It happens first as dollars flow out of the information management ecosystem to fuel the work of often expensive, deliberately temporary players that promise to provide a system-level view and unbiased recommendations. It happens again, as those consultants build their own maps and recommendations only by drawing heavily on the time and knowledge of the very specialists they are shadowing — asking them to explain what they do, who the key voices are, who is and isn’t connected, and where funding is most needed. And then it happens again as the specialists don’t spare attention for funding calls and proposal deadlines because they are spending their time doing the work. 

An unresolved strategic question

Underneath the funding gaps lies a more fundamental question that the field has not yet answered: what is today’s data rescue and resilience effort actually building toward? There is a significant difference between building survival infrastructure to keep specific repositories and datasets alive through the current period of threat, versus building perpetuity infrastructure – systems and services that repositories and datasets can plug into for long-term, stable access regardless of political conditions or other threats. Both are needed. Neither is being resourced with that distinction in mind.

Key, instructive models and seasoned players, ones that have actively solved this problem for adjacent challenges, seem to be missing from the current networks that are assembling: for example, CLOCKSS and Portico for journal content, DataCite for research data identifiers, Crossref for publication metadata, OpenAlex as an open catalog of scholarly works, and Digital Preservation Coalition and National Digital Stewardship Alliance for community standards. Each of these longstanding and well-embedded infrastructure elements has succeeded by defining a function, building infrastructure to serve it collectively, and creating a governance and funding model that distributes the cost across many stakeholders. None of them emerged from crisis response; all of them were built in periods of relative stability. Many of them could provide strong, grounded perspective and leadership to help tie today’s activity to years of preservation practice. 

The current moment is not stable, but it is not too early to ask what long-term preservation and access infrastructure for federal data looks like, and to begin directing some portion of crisis-response investment toward building it. 

What this means

The data resilience field is doing to data rescue what scholarly communications has done to open infrastructure for thirty years: funding the visible and easy-to-narrate activities while underinvesting in the operational and technical layer that makes those activities durable. Coordination without tools is a map without roads. Rescue without perpetuity infrastructure is a temporary holding action, not a long-term solution.

Three specific redirections could materially improve the current investment picture. 

• First, dedicated funding for the tools layer, including web archiving software, ingest pipelines, metadata infrastructure, and identifier systems needs to appear in funding portfolios alongside rescue and coordination projects. Both types of funding are needed, and neither can be substituted for the other. 
• Second, structured knowledge exchange with communities that have done this work before, e.g., the digital preservation community, the 2016-2017 data rescue efforts, SUCHO and other more recent engagement, should itself be an active, funded transfer-of-knowledge activity. Where are there tools, models, and examples that we might build on, and who can quickly help us identify their value? Funding these groups to do this research could vastly improve our readiness to act in the near future.
• Third, instead of funding external groups to tell us what our experts know, let’s fund the lynchpins that are already in place and doing the work, and let’s free up some of their time to help surface and advance the models they are already successfully using. Too many of the organizations best positioned to lead this work, including the established data stewards and the coordination hubs, are operating on project-to-project funding with no runway for strategic planning. Shoring up those existing human and organizational network agents so that they can shift from crisis response to longer-term strategizing is likely to be the best investment we can make. 

The urgency of this moment is real, and so is the opportunity it offers. The infrastructure we build in response to this crisis will shape what is possible for data access and preservation for a generation. Building it on the same underfunded, fragmented, misaligned model that has characterized open infrastructure for decades would be a preventable mistake.

IOI is actively mapping the data resilience landscape and tracking investment trends across the ecosystem, and as we do so, we are grateful for the help and perspective of many other groups, including American Geophysical Union, Center for Open Science, Data Rescue Network, Internet Archive, Public Environmental Data Partners, and The Data Foundation. Please contact katherine@investinopen.org to contribute to or build on this analysis.

Duration: Ongoing
Team: Katherine Skinner, Kaitlin Thaney, Lauren Collister, Madelyn Waterbury
Funders: Wellcome, Digital Science, Arcadia, Karger Publishers Foundation, Kahle Austin Foundation, EBSCO, Lyrasis
Skillset: Business Development Advisory, Capacity Building, Collective Funding

Latin America has built one of the most successful federated open science networks in the world. LA Referencia's aggregation and discovery platform connects national repositories across Latin America and Spain, making research from more than 500 institutions easier to share, preserve, and discover. It's a model of what regional coordination can accomplish, and the region is ready to take it further.

LA Referencia's proposal to the IOI Fund for Network Adoption stood out from more than 100 applications across 22 countries for the scale of its ambition, its track record of effective delivery, and its clear vision for the future.

What the project will do

LA Referencia will extend its platform to an additional 10 Latin American countries, creating a more inclusive and equitable regional network and bringing national science agencies and their repositories into a shared infrastructure.

The platform itself is getting a major upgrade. An AI-powered multilingual semantic search system will help researchers discover and connect work across languages, supported by metadata enrichment that opens the door to new metrics and more transparent approaches to research assessment throughout the region. The project will publish close to 6 million enriched metadata records through this system.

Underneath it all, LA Referencia is building a decentralized persistent identifier system and long-term metadata preservation based on blockchain technology and already running at IBICT in Brazil. A regional Dataverse repository for orphan datasets, along with expanded training and translated documentation, rounds out the work.

Where IOI comes in

Infrastructure at this scale thrives on more than good engineering. It needs a story the world can hear, a business model built to last, and connections to the people who can help it grow. That's where our support is focused.

We're helping LA Referencia organize and amplify its communications so the project's progress reaches funders, institutions, and peer networks across Latin America and beyond. We're also providing project monitoring to support a shared picture of what's working and where momentum is building.

Alongside that day-to-day work, we're researching the strategic questions that will shape the platform's future, including licensing models and business model options for the long term, and connecting LA Referencia with other projects and partners across our network.

What’s next?

By the end of the project, LA Referencia's platform will reach further than ever before, with a multilingual discovery layer that helps researchers find work unbound by linguistic barriers and a decentralized preservation network that keeps research outputs permanently accessible and verifiable. Technical capacity grows right along with it.

The larger outcome is strengthened digital sovereignty for Latin America: research infrastructure that the region's own institutions build, govern, and sustain, with scholarship in Spanish and Portuguese as discoverable as work published anywhere else in the world.

Have a vision for your region?

Some of the most exciting open infrastructure work happens when institutions pool their strengths across borders. If your network has an ambitious idea and needs the funding, business planning, or connections to bring it to life, we want to hear about it. Get in touch to learn how IOI's Coordinated Funding efforts pair investment with hands-on support.

Outputs:

• Empowering networks, advancing openness: Invest in Open Infrastructure announces inaugural grantees of the IOI Fund for Network Adoption

Duration: 2024 – present
Team: Chrys Wu, Lauren Collister, Emmy Tsang, Sarah Lippincott, Katherine Skinner, Gail Steinhart
Funders: Mellon Foundation, Arcadia, Digital Infrastructure Insights Fund, National Science Foundation, Sustaining Circle members
Skillset: Ecosystem Intelligence, Landscape Analysis

Overview

The open infrastructure landscape is large and growing but the information needed to navigate it is scattered, inconsistent and held by no neutral party with a view across the whole. There are platforms, standards, repositories, and persistent identifier services doing important work, but they can be difficult to discover and even harder to evaluate side by side. For funders, that made it difficult to direct resources well. For adopters, it made decision-making slower and riskier than it needed to be. And for the infrastructures themselves, it made visibility a persistent challenge.

IOI, funded by the Mellon Foundation, set out to build something practical: a public tool that gave both funders and adopters the clear, verified, comparable information they needed.

What we did

Before building anything, we needed to understand what funders and adopters actually needed. IOI’s position at the intersection of both stakeholder groups meant that we knew who to talk to and what the real questions were. We spoke with more than 40 institutional budget holders and 26 infrastructure service providers in one-on-one meetings, focus groups, and user testing sessions. 

From there, we built Infra Finder as a public-facing product: a searchable, filterable database of open infrastructures that could grow with input from the community. Each entry covers the areas that matter most to evaluators: technical attributes, governance, community engagement, key policies, and funding models — each area a different facet of what it means to be "open".

Getting the information right required more than a one-time data collection effort. We also built the workflows to source, validate, and keep that information current, working directly with infrastructure providers to make sure what users find in the tool is accurate and up to date. A comparison view lets users look at up to four services side by side, with links to deeper detail where they need it.

Outcomes

Infra Finder launched with coverage of 57 infrastructure services and has since grown to more than 130 thanks to community input. 

For funders, it offers a clearer view of the landscape than existed before, with enough detail to inform real investment decisions. For adopters, it shortens the research process and reduces the risk of choosing infrastructure that turns out to be a poor fit.

For the infrastructures themselves, it creates a new kind of visibility. As Niels Stern, Managing Director of the OAPEN Foundation, put it: "Participating in this tool provides us with the opportunity to disseminate a lot of information in a very concise way and be more transparent about what we do. This is useful because we can share our entries with institutions, funders, and anyone who wants to work with us or is interested in learning more about our work."

That transparency matters. Open infrastructure often does essential work quietly, without the visibility that helps funders find it or institutions trust it. Infra Finder was built on the premise that better information, in one place, for everyone, changes that dynamic.

Impact by the numbers

In the two years since its inception, Infra Finder has become a go-to resource for open infrastructure diligence in the scholarly community. The site has received over 30,000+ visits, 40,000+ unique pageviews from 130 countries globally since launch. Of these numbers, close to 7,000 returning visits (23% of total visits), highlighting its value as a trusted resource.

Finding your way through the landscape 

The open infrastructure ecosystem does not stand still. Platforms emerge, mature, merge, and sometimes close. Funding models shift. The expectations placed on infrastructure by institutions, by policy, and by researchers keep growing. Keeping track of what exists, what it does, and whether it is still the right fit for your needs is genuinely hard work.

If you're trying to make better decisions about where to invest or what to adopt, Infra Finder is a great place to start. And if you need support thinking through what those decisions actually require, we would love to chat!

Duration: Ongoing
Team: Katherine Skinner, Kaitlin Thaney, Jerry Sellanga, Lauren Collister
Funders: Wellcome, Digital Science, Arcadia, Karger Publishers Foundation, Kahle Austin Foundation, EBSCO, Lyrasis
Skillset: Business Development Advisory, Capacity Building, Collective Funding

Overview

Research from across Africa is growing in volume, ambition, and relevance, and the world benefits every time it becomes easier to find. Open institutional repositories make that happen, keeping scholarship discoverable, citable, and preserved for the long term.

UbuntuNet Alliance is well positioned to build that repository layer at regional scale. Its 16 National Research and Education Networks (NRENs) serve roughly 846 institutions across Eastern and Southern Africa, and they hold something no other provider can replicate: deep trust within their research communities. The Alliance's proposal to the IOI Fund for Network Adoption stood out from more than 100 applications across 22 countries for its ambition and its practicality in equal measure.

What the project will do

Its member NRENs, with coordination by UbuntuNet Alliance, will roll out DSpace as a managed cloud service, giving institutions repository services on infrastructure they already know and trust. More than 560 librarians will train as repository managers and metadata librarians, creating or strengthening over 400 open institutional repositories across the region.

The project will also build out AfricArXiv as a pan-African aggregator for open research metadata, and train more than 450 data management champions to support a target of over 960 curated open datasets that can be cited, discovered, and reused worldwide.

Where IOI comes in

The strongest open infrastructure pairs good funding with a solid business model, sound governance, and a community that owns it. That's why the Fund was designed to deliver both. Every grant comes with hands-on strategic support, and that's where our work with UbuntuNet Alliance is focused.

• Business planning that fits the region: We're running business planning workshops and training with NRENs across Eastern and Southern Africa, drawing on IOI's experience in nonprofit revenue models, collective funding approaches, and funder diversification. The goal is for each network to develop a sustainable model that serves its library and researcher communities well beyond the grant period.
• Listening before designing: We're conducting interviews across the region to build a deeper understanding of local needs and trends in repository development, so that business models and services reflect what institutions actually want rather than assumptions about what they need.
• Connecting to the global community: DSpace is a mature open-source platform with an active worldwide community and established governance. We're connecting UbuntuNet Alliance directly into that community, so the region's voice shapes the platform's direction and UbuntuNet Alliance's teams can draw on global expertise.

What’s next?

By the end of the project, the region will have a trusted metadata aggregator of its own, hundreds of new and strengthened repositories, and NRENs equipped with the business models to sustain these services for the long haul. The work aligns with the African Union's Agenda 2063 and the UNESCO Recommendation on Open Science, and it strengthens South-South research collaboration in the process.

The larger outcome is the one that matters most: African scholarship that is visible, valued, and verifiable, available on infrastructure that the region's own institutions run and govern.

Building something at network scale?

Networks and consortia hold a kind of trust that makes open infrastructure adoption move faster and stick. If your network is thinking about how to scale repository services, strengthen sustainability, or connect with funders who back this work, we would love to talk. Reach out to learn more about the IOI Fund for Network Adoption and what integrated funding and strategic support can look like.

Outputs:

• Empowering networks, advancing openness: Invest in Open Infrastructure announces inaugural grantees of the IOI Fund for Network Adoption

<< Back to Our Work

Why this change?

People, and their privacy, safety and rights, remain at the heart of IOI's work. As our organization and operations have grown, we've revisited the data we collect, how we store it, and how we work with it, to make sure our practices stay robust and continue to serve our community and everyone we interact with. This update reflects that review.

What should I do?

Please read the current privacy policy and note your rights under it. You can find the previous version via the Wayback Machine.

If you have any questions, email us at data-request [at] investinopen [dot] org.

We review this policy on a regular basis.

Sharon Leon is Co-CEO of the Corporation for Digital Scholarship. She has worked across Digital Scholar's software projects since their inception, originally as faculty at the Roy Rosenzweig Center for History and New Media at George Mason University and then at Michigan State University.

A humanistic approach to open infrastructure sustainability beyond grant-funded beginnings

Digital Scholar was founded in 2009, which surprises people when Leon mentions it. The original purpose was straightforward: find a way to sustain core open source software, specifically Omeka and Zotero, so that they could continue to exist without depending indefinitely on grants. The founders' approach was shaped by who they were. Leon and her colleagues were trained as historians, and that perspective shaped how they thought about the people who would use their tools, even as that user base eventually extended far beyond digital humanities or the academy.

"Zotero has 17 million users around the world," Leon notes. "They're certainly not all humanities scholars. They're in the legal field, the sciences, the corporate world, all over. Our perspective shows us that the users of our software are not necessarily folks who can build their own tools. They're not necessarily super technically adept, though many are." Digital Scholar's insistence on interfaces and documentation that make research tools genuinely accessible — resisting the temptation to assume users can just write a script when they encounter a problem — flows directly from that founding orientation. "What we are keeping in mind are the humans doing the digital work, in addition to the humanistic subject matter," Leon says.

Their work has expanded beyond those two initial projects. Digital Scholar now stewards five core projects: the original two (Omeka and Zotero), with PressForward, Tropy, and Sourcery joining through the years. The staff has grown as well, expanding from a very lean operation to a team of roughly 35 people.

Explore Digital Scholar’s entries on IOI’s Infra Finder. 

Building sustainability through services

The sustainability solution the founders landed on was to offer services alongside the software in response to user needs. "People are willing to pay a little bit for storage, or for hosting, or some other helpful service," Leon explains. "And that, in turn, allows us to pay the project teams to keep the software going." 

Central to this approach is a core question: what do people find valuable enough that they're willing to pay for it? The signals come from users themselves, sourced from the Digital Scholar team monitoring closely requests that surface in forums, discussion posts, and GitHub issues; these requests reveal where genuine needs aren't yet being met. "In the case of Omeka, it was a clear sense that not all of our users had the time or interest in setting up a server themselves. We thought, wouldn't it be easier if they could just press a button and start?" That signal led to a popular hosted Omeka service. Similarly, Digital Scholar offers paid storage upgrades for Zotero, making it possible for individuals and groups to sync and share materials.

The organization's origin story includes an unusual institutional detail that is fundamental to how it operates in order to sell these services. Digital Scholar is not a 501(c)(3) like many nonprofit organizations in the open infrastructure space in the United States. Rather, it is a nonprofit, nonstock corporation in the Commonwealth of Virginia, meaning the state confers nonprofit status, but the organization carries no IRS tax-exempt designation. "We don't have a membership model," Leon says. "We actually sell services. And selling services is a different kind of model; we pay income tax, we collect sales and remit tax, we do all of those things that a regular business would. But we have no owners, no shareholders, and no way to do anything with the income except support our core software projects and support the field."

This structure also lets Digital Scholar play a direct role in strengthening the broader research ecosystem. Because its business purpose includes support for open source software and open access work in digital scholarship and cultural heritage, Digital Scholar can make direct charitable donations to other nonprofits or it can directly underwrite projects at other institutions (such as its support of the Digital Humanities Now news outlet at Centers for Digital Scholarship at the Northeastern University Library). This model also lets Digital Scholar explore ways to contribute in-kind resources and staff time, or offer access to its business systems and financial infrastructure. This flexibility has opened the door to a new kind of work.

Extending the model: fiscal support and organizational mentorship

In recent years, Digital Scholar has begun offering a different kind of support to other projects in the open infrastructure ecosystem. Known within the organization as Slipstream, it is not yet a formal, open program, but two current relationships illustrate the shape of what Digital Scholar is developing.

The first is with Mukurtu CMS, a cultural heritage platform that has been around nearly as long as Omeka. Mukurtu began offering managed hosting to generate revenue outside of grant funding, and turned to Digital Scholar to handle the operational side: purchasing, server procurement, invoicing, and contracting. "We have that business infrastructure that we can share with them," Leon says. "Net income from running all of those systems get transmitted back to the project to support it." For a mature project like Mukurtu that doesn't need intensive guidance, the relationship is largely practical; Leon describes it as a shared business layer that eliminates duplication of administrative effort.

The relationship with RightsStatements.org is more intensive. RightsStatements.org is a critical piece of rights infrastructure for open access to digital cultural heritage, originally born from collaboration among major aggregators including DPLA and Europeana. But over time, through COVID and shifting institutional priorities, its governance structures had become fragile. The project's Interim Steering Committee was dedicated but lacked an independent legal entity and the organizational capacity to sustain the work.

When RightsStatements.org put out a call for potential new institutional homes, Digital Scholar responded. "We realized how important they are to the global working of understanding access to digital cultural heritage materials," Leon says. The result is an initial three-year relationship in which Digital Scholar is working with the Interim Steering Committee to rebuild governance, creating structures that allow the community to participate in decisions about the statements, their updates, and their translations. The goal is, ultimately, to set the project up for success as its own independent organization.

The two relationships represent distinct points on a spectrum: with Mukurtu, they are offering shared services building on existing expertise, while with RightsStatements.org, they are participating in an intense hands-on project of building a governance framework from the ground up. Through these two projects, Digital Scholar is discovering how much capacity it actually has for this kind of support; that knowledge will shape its offerings going forward.

Planning infrastructure for the future

Open source infrastructure, Leon observes, faces a particular challenge: the software can persist indefinitely, but to do so, the sustainability of the people who work on it must be considered. "Infrastructure may not be coming to the end of its life," she says. "But the humans whose careers have been centered around supporting it — at some point, those people have to be allowed to move on, whether to retirement or to something else. But infrastructure is infrastructure, and people rely on it."

Her advice: think about sustainability from the beginning. Not as an afterthought, not as a future grant application, but as a foundational design question. There are still too many pieces of the infrastructure the research community relies on that haven't accounted for the long-term. 

For their part, Digital Scholar is entering a new strategic planning cycle, and the agenda ahead reflects the same user-driven logic that has guided the organization since its founding. Leon shared a few developments on the horizon that users can look forward to. 

PressForward, a WordPress plugin for aggregating and curating scholarly gray literature, is in the early stages of being extended into a standalone web service; fewer people run WordPress blogs, and content now moves through Substacks, newsletters, and static site generators. "The world of scholarly communications has changed, and the tool needs to meet users where they are," she says. The team is also thinking carefully about preservation and portability: connectors for ArchivesSpace and Archivematica extend Omeka’s reach into the archival ecosystem, and static site exporters now allow Omeka sites to migrate to Hugo for retirement and/or preservation. On artificial intelligence, Digital Scholar is watching carefully and beginning to support some applications through opt-in plugins — but nothing by default. For instance, Tropy is developing a computer-generated transcription service for research materials. "The users will guide us on the uses they feel comfortable with," Leon says.

An example of the new handwriting transcription feature in Tropy.

At the same time that they are cultivating their own tools and communities, Digital Scholar is also actively growing its role as a support and mentorship organization for the broader open infrastructure ecosystem. What Digital Scholar offers (business infrastructure, governance support, hard-won institutional knowledge) exists in service of a field that still has too many projects struggling to survive the gap between initial grant funding and genuine sustainability. 

Building on 15 years of foundational support for critical open infrastructure tools, Digital Scholar is responding to the needs of researchers, creators, and stewards to strengthen the research ecosystem from the ground up. 

The Corporation for Digital Scholarship (Digital Scholar) stewards Omeka, Zotero, PressForward, Tropy, and Sourcery, and provides organizational support to RightsStatements.org and Mukurtu. Learn more at digitalscholar.org and visit their entries in Infra Finder.

One thread that we found was that the extant literature nearly uniformly advocates commons-based approaches as alternatives or complements to market-based strategies for open curated collections. The commons approach emphasizes collective coordination, social norm cultivation, and institutional trust-building rather than monetization of individual collections. Unlike licensing deals struck between AI firms and individual publishers, commons-based approaches seek to preserve the open, shared character of knowledge resources while ensuring that the entities profiting most from them contribute meaningfully in return. This perspective positions small collections not as individual vendors but as stewards of interconnected knowledge resources requiring collective protection.

The success of a commons-based approach assumes that AI companies do have incentives to engage with open collections stewards—incentives not predicated solely on legal risk. These incentives include: 

Ensuring high-quality data sources remain online. Excessive bot traffic and scraping threatens to push some small collections offline, as many lack the resources to "continue adding more servers, deploying more sophisticated firewalls, and hiring more operations engineers in perpetuity" (Weinberg, 2025; Grant, 2025). If key data sources go dark, AI companies lose access to the very content that makes their models useful.

Encouraging competition and discouraging consolidation. Investment in the commons enables everyone, not only the wealthiest corporations, to build and refine models that lead to innovation. Even larger companies share a broad interest in encouraging innovation in the sector that they can benefit from in the future.  In its announcement of its support for Harvard’s Institutional Data Initiative, Microsoft cited the motivation to grow “a vibrant, competitive AI economy” by expanding access to the data resources needed to build LLMs (Davis, 2024). Openly licensed datasets can encourage competition and offer smaller players a way in. Adoption of the Model Concept Protocol (or MCP, initially developed by Anthropic and later donated to the Linux Foundation) by the major AI market players is an example of industry-wide cooperation in this vein.

Retain scraping access. The relationship between AI companies and the broader web is already showing signs of strain, and the consequences of ignoring this dynamic are already visible. A closing off of the web in response to AI crawlers, especially through blunt approaches that do not distinguish them from other machines, is affecting crawling for legitimate and widely accepted purposes, such as archiving and research. As of December 2025, around 5.6M websites had blocked OpenAI's GPTBot, a nearly 70% increase over the previous six months (Claburn, 2025). As scraping restrictions increase and more organizations adopt brute force approaches to thwarting bots, AI companies risk losing access to important sources of high-quality, novel training data. Investing in the health of the commons, and in norms that distinguish reasonable use from indiscriminate extraction, can help arrest this trend.

Sustain high-quality, diverse training data. Well-stewarded commons provide not just volume but also the metadata, documentation, and quality control that make training data more valuable. Companies may see value in building and sustaining resources that provide them with access to high-value, unique, or novel datasets. Some analyses have speculated that the open web will become polluted by low-quality, machine-generated content, making curated collections increasingly valuable data sources. Noroozian et al. (2025) write that AI model developers should have a vested interest in making curated collections data "identifiable, visible, and discoverable" in order to avoid 'model collapse' or increasingly more repetitive, biased, and less capable AI" caused by the growing presence of synthetic data across the web. Looking further ahead, Woahn (2026) predicts that "The next improvements in model capability will come from: highly specialized domain corpora; well-structured technical datasets; targeted refreshes rather than massive new ingestions; data with deep internal organization, not broad volume."

Foster ongoing human contributions to the open web. The commons is sustained by the continuous labour and ingenuity of human creators. Without new approaches for providing permission, credit, and compensation, these creators have diminishing incentives to openly share their work, and AI models lose access to original content (Chan et al., 2023, Huang & Siddarth, 2023). Borgman and Groth (2025) argue that scholars participate in a gifting economy in which they volunteer labour (such as sharing data) "with the expectation that these gifts create indebtedness, encourage reciprocity, and enhance reputations." To build trust among scholars and collections stewards, AI companies may need to more visibly and concretely adopt the norms of a gifting economy, for example, by ensuring proper attribution.

Create a positive public image and consumer trust. Beyond practical considerations, AI model developers may have a reputational incentive to demonstrate a commitment to "ethical" or "responsible" AI, including appropriate data-harvesting practices. The non-profit Fairly Trained, for example, was launched to certify AI model developers and products that adhere to standards for their training data (Knibbs, 2024). They also have an interest in providing consumers with reliable information from robust sources to increase adoption and engagement with their platforms.

Mitigate regulatory and legal risks. Finally, the legal landscape surrounding AI and data use remains in flux. Depending on the outcomes of several lawsuits and pending legislation, AI model developers may need to fundamentally alter how they harvest data. If they cannot rely on fair use justifications for scraping copyrighted data, for example, they will be increasingly reliant on openly licensed and public domain data. The strongest incentive for change could be future government policy that regulates the use of openly available data, for example, by strengthening creator opt-outs.

Despite these shared incentives, the challenge lies in bridging the gap between curated collections stewards and AI companies and developing the sociotechnical infrastructure needed to facilitate cross-sector engagement. Trust between commons communities and AI companies is severely eroded. Open source developer communities have expressed "deep frustration with what they view as AI companies' predatory behaviour toward open source infrastructure," undermining the relationship-building these approaches require (Edwards, 2025). Philosophically, there's tension between ideals of openness and the need for protection. The "open with thoughtfulness" paradigm (Metz, 2025) requires continuous judgment calls that may fragment the commons into incompatible governance zones.

Addressing these challenges will require deliberate effort on multiple fronts. The commons needs norms, governance frameworks, and contribution models developed with input from a range of stakeholders, including AI companies and technology platforms, as well as researchers, creators, and open curated collections stewards. The BRIDGE project is ongoing and represents IOI’s current efforts to identify and foster relationships between these currently disparate groups, and to pilot new approaches to encourage reciprocity that benefits all stakeholders in the AI economy. 

This post is an excerpt from our longer work, "Sustaining the Commons in the AI Economy: A Landscape Scan of Challenges and Strategies for Bridging AI Companies and Open Curated Collections." 

As part of our project release, IOI’s Chrys Wu interviewed the report’s authors (Sarah Lippincott, Lauren Collister, and Katherine Skinner) about their perspectives on the main takeaways from this report. This interview is now available to watch or listen to. 

We hope you enjoy this interview with the authors, and we would like to hear your take-aways as well. What parts of the report most resonate with you? Send us a message at research @ investinopen dot org, or respond to our LinkedIn post to share your view. 

Read about the full report at Sustaining the Commons in the AI Economy: A Landscape Scan of Challenges and Strategies for Bridging AI Companies and Open Curated Collections.

Curated collections, such as digital archives, open access journals, scientific data repositories, preprint servers, and knowledge graphs, form part of the digital commons: the shared pool of open resources made freely accessible online. Used regularly by researchers, the public, and commercial entities, this digital commons benefits every sector of society. Maintained largely by academic institutions, nonprofits, governments, and volunteer communities, these curated collections represent a public good built on decades of labour, public and private funding, and an ethos of open knowledge sharing. Their sustainability is inseparable from the sustainability of open science and democratic access to information.

That infrastructure is now under strain. The rapid expansion of AI development has turned open curated collections into an increasingly valuable source of training data. Automated bots now generate traffic that, in some cases, exceeds human visits, overwhelming servers, inflating bandwidth costs, and triggering outages. Meanwhile, a surge of AI-generated content submissions threatens to overwhelm editorial and curation workflows at repositories that rely on community contributions. 

To understand this landscape, key themes explored in the report include:

• The strain on infrastructure: How AI bot traffic is overwhelming servers, inflating costs, and triggering service disruptions at open curated collections.
• The limitations of current approaches: Why technical, legal, and market-based mechanisms face significant challenges in protecting the commons at scale.
• The risks of defensive restrictions: How access controls intended to protect collections may paradoxically accelerate data consolidation among well-resourced corporations.
• The tension of voluntary compliance: Many current approaches rely on voluntary compliance, with legal repercussions as enforcement mechanisms; what else could be done to encourage cooperative behaviour? Wrestling with whether enlightened self-interest can succeed where legal and technical frameworks have fallen short.
• The possibilities of the commons: AI companies and collection stewards both have much to gain from exploring the commons as a shared space for investment and cooperation. 

The report points toward a promising, if demanding, path: commons-based governance grounded in reciprocal norms and shared interests across sectors. AI companies have concrete reasons to want the commons to survive. The loss of reliable data sources reduces the quality and diversity of training data; an increasingly walled-off web raises legal and regulatory risks; growing public frustration with extractive practices creates reputational pressure. Investment in the commons, properly framed, is investment in the quality of AI itself.

Yet a central tension remains unresolved. Whether enlightened self-interest will prove more effective than the legal and technical mechanisms that have already fallen short is an open question. To answer it requires engaging the curators and consumers of open collections as stewards of the digital commons, co-creating partnership models that could align open knowledge strategies with commercial demand.

We're eager to hear from interested parties and continue the conversation. If you'd like to be part of further discussions on this topic, please contact us at research [at] investinopen [dot] org.

In response, Proyecto ARPHAI sought to develop a project to accelerate and promote the ethical use of digitized health records nationwide. Led by CIECTI (the Interdisciplinary Centre for Science, Technology, and Innovation Studies), ARPHAI researches tools based on artificial intelligence and data science that can be applied to electronic medical records. By strengthening both the analytical value of health records and the technical and human capacity to work with health data, ARPHAI contributes to earlier detection of epidemic outbreaks and supports more timely, preventive public health decision-making.

Collaboration with IOI

In 2023, IOI opened the call for proposals for the Open Infrastructure Fund, which aimed to support the development of open research infrastructure services to strengthen sustainability, resilience, and adoption of open infrastructure. Proyecto ARPHAI was selected as one of the eight grantees of the fund, receiving funding of around US$18,000 to develop and sustain the infrastructure for processing and storing sensitive data from electronic health records in Argentina.

Implementation and Impact

The one-and-a-half-year project, which commenced in March 2024, was implemented in collaboration with the Centro de Computación de Alto Desempeño (CCAD), a high-performance computing facility at the Universidad Nacional de Córdoba (UNC), Argentina. A key deliverable was the acquisition of a high-capacity 'fatnode' server — affectionately nicknamed Gordito (Spanish for 'chubby') — equipped with 1 terabyte of RAM and support for 8 next-generation GPUs. The hardware provides the core computing infrastructure required to host and process electronic health records in a secure environment. The servers will support data storage, analysis, and the development of artificial intelligence models, enabling researchers to work with sensitive health data while maintaining responsible data governance. The server runs within CCAD's Kubernetes Cluster, hosting both Project Jupyter tools for interactive computing and Ollama for running large language models. The additional computing capacity has also benefited the broader teaching and research community at CCAD, extending its value well beyond health data applications.

“This server is a key piece for the JupyterHub at CCAD, which was recently inaugurated, because it is the one that makes it possible to run notebooks with one terabyte of RAM and up to four GPUs, providing computing power for the largest projects. The facility has also been a big boost for our wider research community at CCAD as we now have more computing power available even for researchers in other areas beyond healthcare,” commented Nicolás Wolovick, Director of CCAD.

To ensure that sensitive data stored on national servers yields actionable policy insights, the ARPHAI team organized two training sessions for government health officials in May and July 2025. These interactive sessions gave officials a solid foundation for working with sensitive electronic health data and for leveraging insights derived for the public good. Extensive documentation and training materials in Spanish were developed and made freely available to the Argentine research community under CC-BY licences, hosted on both the ARPHAI repository and Zenodo.

The importance of using local servers when dealing with a population’s health data is key to ensuring digital sovereignty and the protection of personal data, and strengthens states’ ability to manage data autonomously and in line with public interest objectives.

“It was a great pleasure for us to share our experience and help pave the way for those who face the daily challenge of working with sensitive data. It was doubly rewarding, not only because of the interest and feedback generated by the training, but also because it was an achievement in itself to get people so busy with the demands of daily management to take a moment to reflect on and analyze the impact of their work and their approach to it. Today, in a context where science funding faces significant challenges (particularly in countries like Argentina), these kinds of initiatives are especially valuable, as they contribute to sustaining and strengthening established research groups with demonstrated expertise and ongoing activity,” remarked Sabrina Lopez from Proyecto ARPHAI.

Lessons Learnt

Technical investment must be matched by human investment. The project design accounted not only for hardware but also for the capacity building needed to ensure long-term sustainability. Projects that focus solely on the technical side often face sustainability challenges. Community buy-in and well-trained personnel are equally critical to success.

The importance of partnerships in funding and sustaining open infrastructure. In recent years, funding for the science and education sector in Argentina has been cut. This, in turn, has created a massive financial deficit. From the project, we can see exemplary collaboration between different stakeholders (universities, research organizations, infrastructure services, and non-profits) to work together for the common good by pooling resources, which needs to be emulated.

ARPHAI's vision was realized through its partnership with CCAD, not in isolation. CCAD hosts the project’s infrastructure and data and supports the development of an active community of practice. CCAD had supported ARPHAI's research for two years before the IOI funding was secured — a testament to the importance of cultivating strong, mutually beneficial relationships in advancing societal impact.

Embed, don't impose. Infrastructure investments are more likely to see organic adoption when they flow through organizations and networks already embedded in their communities. These organizations bring the trust and contextual understanding that funders can't manufacture — they know what to build, for whom, and why it will actually get used.

Conclusion

Proyecto ARPHAI demonstrates what becomes possible when technical ambition is matched by the right partnerships, development of human capacity, and a commitment to openness. The challenge now is to scale this model to ensure that other sectors beyond healthcare can leverage open infrastructure for broader, sustained impact.

IOI's Director of Development, Emma Green led the engagement, working with 2i2c's Business Development Manager, Jim Colliander, and supported by Chrys Wu, Solutions Strategist at IOI. Jim reflected on what the partnership looked like in practice, what it made possible, and the key takeaways from the engagement. Below are some highlights.

When IOI first began working with 2i2c, what were the most urgent business or growth challenges you were trying to solve, and how were those affecting your ability to operate sustainably?

Jim: The core challenge was straightforward but not easy: we needed a credible path to organizational sustainability. Since 2i2c’s formation in 2020, we have built an initial base of recurring revenue. However, much of that recurring revenue had been secured through founders' networks and warm referrals; a pattern the team recognized as unsustainable. 

At the same time, we were navigating a significant internal transition. In early 2025, we moved from offering a managed JupyterHub service to a membership model, partly to make the full range of value we provide more explicit. That shift raised harder questions about what kind of organization we actually wanted to be. Some of the team saw our path to scale as becoming more like a SaaS product. Others felt it required more consultancy-style, bespoke engagement. IOI helped us work through that tension and arrive at a shared definition of what scaling actually means for 2i2c — which turned out to be important for our next steps.

What key assumptions about your role and approach to market, pricing, or go-to-market approach did IOI help you test or rethink, and which of those turned out to be the most important?

Jim: IOI guided 2i2c to take a more hypothesis-driven approach to go-to-market strategy, starting with identifying a key assumption that 2i2c had made: that 2i2c could successfully sell directly to Premier-tier customers through cold outreach. IOI helped us think about go-to-market with improved terminology and a scientific approach that centers user needs through user discovery processes.

What we discovered surprised us. Customers were not only seeking access to hubs. They were looking for genuine engagement with the 2i2c team, connections to other 2i2c communities, and a sense of participation in something larger.

On a personal level, the engagement also shifted my perception of my role. I had been conflating three distinct functions: the Farmer (focused on retention and expansion), the Hunter (focused on net new sales), and the Business Development Lead (focused on strategy and building growth systems). Recognizing which hat I was wearing at any given moment changed how I thought about scaling sales.

How did IOI’s engagement influence conversations/enhance alignment between business development, product, and engineering within 2i2c?

Jim: We discovered that we had real internal misalignment — not just on strategic direction, but on how commitments from the sales side were being translated into scoped work for product and engineering, and on how customer feedback was feeding back into what we built.

IOI recommended a recurring customer insights meeting that brings together business development, product, and delivery teams to create a structured cross-functional feedback loop. This has been implemented and we are already seeing improvements in the business-to-product feedback loop within 2i2c. IOI introduced and documented sales ceremonies. We now have a daily Business Development standup, a bi-weekly business strategy meeting, and a weekly engagement meeting. IOI’s Emma joined some of 2i2c’s BD meetings near the end of the engagement to provide real-time coaching and follow-up guidance.

Can you describe a concrete change in how 2i2c now approaches the BD role, sales, partnerships, or product design that grew directly out of this coaching and advisory work?

Jim: IOI helped us develop clear written role definitions distinguishing the Account Manager function from the Business Development Lead function. That clarity reduced some friction.

We also adopted a SAM/TAM horizon framework — mapping our Serviceable and Total Addressable Markets across a three-horizon planning timeline — which shifted our thinking from quarterly wins to a two-to-three year view of market expansion. That longer lens has been useful.

We built a practice of customer discovery. After IOI coached us on survey design, bias, and validation thresholds, I ran a structured customer survey experiment with eleven respondents. The results shaped our product direction and improved our internal consensus.

Looking ahead, how has this work reshaped your confidence to contribute to and drive and what ‘scaling’ realistically means for 2i2c as an open infrastructure organization?

Jim:  The most clarifying moment in the entire engagement was probably the reframing of what "scale" actually means for an organization like ours. Our first target isn't AWS-level growth — it's building repeatable, legible processes for five to ten customers of the same type. That sounds modest, but getting there requires exactly the kind of disciplined system-building we needed..

Emma's coaching shifted something more fundamental too. My focus used to be on getting things done; now it's on building evidence-based systems and knowledge that the whole team can work from. That's a different job in some important ways. The 2i2c team is more focused now on learning and building strong internal habits than on task completion — and I think that's the right foundation for where we want to go.

This post is part of our “Infrastructure Showcase” series. To stay updated on posts from this series and more from Invest in Open Infrastructure, please sign up for our newsletter. Interested in IOI’s strategic consulting services to further research infrastructure health, sustainability, and growth on these topics or related areas of work? Get in touch!