CKAN at the United Nations: Why Digital Public Infrastructure Needs Open Source

In June 2025, the United Nations convened technologists, policymakers, researchers, and civil society leaders from around the globe for UN Open Source Week — a milestone event focused on building trustworthy Digital Public Infrastructure (DPI) using open technologies.

Among the participants were two voices from the CKAN ecosystem: Joel Natividad, co-founder of datHere and long-time CKAN contributor, and Dr. Nora Mattern, a scholar at the University of Pittsburgh exploring libraries, information systems, and public interest technology. Both were there representing the NSF-funded CKAN POSE project, aimed at strengthening CKAN as a foundation for open, participatory, and sovereign data infrastructure.

After the event, I sat down virtually with Joel and Nora to reflect on their experiences — and to explore how the insights from UN Open Source Week connect to CKAN’s evolving role in the global open data landscape. What follows is a deep, honest, and wide-ranging conversation about trust, community, interoperability, and the very nature of digital public goods.

First impressions from UN Open Source Week

To start, could you share your overall impressions of UN Open Source Week? What made the event particularly significant for you both?

Nora: In the lead-up to attending the UN Open Source Week, I spent time learning more about the work of the UN Office for Digital and Emerging Technologies, the organizer of the meeting.

It was useful context and helped me appreciate the commitment to open that the UN holds.

For example, I learned about the Global Digital Compact, now stewarded by the Office for Digital and Emerging Technologies. In addition to articulated goals related to closing the digital divide and building digital literacy skills and capacities, the Compact commits to promoting, developing, and maintaining open standards and software.

I saw a deep relationship between the CKAN community’s recognition of open software as in the public interest and the work of the UN Office for Digital and Emerging Technologies, including in the Compact.

At the meeting itself, I was struck by the representation in the room – from geographic to sector representation.

Since I’ve arrived to the open source and CKAN community from my engagement with open community data initiatives and libraries, I had a lot to learn from the broader discussions.

For example, I did not understand the reliance that the corporate sector has on open source and, in turn, the ways that companies contribute to open source development and maintenance (for example, a member of the Mercedes-Benz team shared the company’s use of and contributions to open source).

I was able to better understand how organizations are responding to their commitment to open source with the development of “open source program offices” (or OSPOs) and the type of roles that they play.

Joel: It was very gratifying to have had the opportunity to sit amongst folks from around the world - not just technologists; but policymakers, thought leaders, advocates, corporates, philanthropists - civil society at large - all talking about “open source” of all things!

If you told me 35 years ago when I graduated with a Computer Science degree that the UN is hosting an Open Source conference, I wouldn’t have believed you!

But that’s the thing - they were not talking about Open Source in a purely technical sense - they were talking about it as a foundational building block of our digital society.

In those 35 years, computers have infiltrated every aspect of modern life, thanks largely to the wide-scale adoption of the Internet - itself, one of the first global scale technologies enabled by open source.

It has made the world a lot smaller - enabling globalization, automation, and the financialization of the global economy.

Yes, it has made the world richer and lifted billions from destitution, but at what cost - to the environment, to our society, to our well-being?

UN Open Source Week demonstrates that policy-makers around the world now recognize the instrumental role of technology in shaping our shared digital futures, and there should be an effort to create it with more intent - like DPI.

DPI and the CKAN mission

Joel, in one of your Linkedin updates, you quoted Vilas Dhar’s DPI framework: "From values and principles, to code. Code to governance. Code to policy. And policy back to innovation." How do we practically embed trust, transparency, and participation into open-source systems like CKAN to truly earn trust in our digital infrastructure?

Joel: I found myself going back to that LinkedIn post several times a day since UN Open Source week as it really resonated with me.

As I agree with Vilas - “there is no such thing as a neutral technology, and there cannot be such a thing as a neutral technologist.”

We need to design and build things with intent - as the typical Silicon Valley playbook of “move fast and break things” has left us exactly that - a trail of broken things.

Arguably, that is the nature of change - but as technology accelerates, especially with AI, can we afford to follow the same script of unbridled innovation for short-term profit without consideration for its long-term consequences?

With CKAN, I submit that it already follows Vilas’ DPI vision - a system built with intent.

For CKAN was birthed by the Open Knowledge Foundation, itself founded by Rufus Pollock in 2004 based on his Open Knowledge vision.

In pursuit of that vision, Rufus started coding CKAN in 2005 and it has only grown since then.

Open source from the beginning, it has evolved from an open data application to a mature data management platform that can do much more than open data - able to tackle the Digital Exchange use cases called for by DPI.

And with CKAN 3.0 development in earnest, I’m excited by the different experiments and innovations being co-created by the Core team along with thousands of contributors around the world.

In particular, CKAN’s DCAT3 support for FAIR (Findable, Accessible, Interoperable, and Reusable) Data can embed several Trust mechanisms into CKAN. Apart from existing mechanisms like Digital Object Identifiers (DOI) and checksums, there are also several experiments with standards that employ blockchain technology to ensure the authenticity and provenance of datasets.

So yes! CKAN has been earning Trust on both social and technical fronts - with the community with its long heritage of open-source development. On the technical front, with the various Trust mechanisms above.

With the continued guidance and stewardship of CKAN by OKFN and the stewards, I’m hopeful that our POSE project will only further strengthen CKAN’s mission-driven, intentional, innovative open culture and make it a trusted platform for building DPI.

Nora, you’re not just a technologist — you’re someone who studies how systems impact human lives. What resonated most for you at the UN Open Source Week? What surprised you?

Nora: For me personally, the conversations that I had on the elevator, in the coffee queue, and over lunch were the most impactful aspects of what was a rich program.

For example, I met a colleague from Creative Commons and learned about their newly announced CC Signals Project, a designation that data stewards can use to communicate whether data can be used to train AI.

I also met a member of the Software Heritage project and attended a talk by its founder. Software Heritage is a non-profit that collects, preserves, and makes available software code, recognizing the importance of software as a record of digital culture.

These are initiatives that I will connect others to, including the library science students I work with and project collaborators.

CKAN as infrastructure — the “airport” metaphor

Joel, you also shared Miller Abel's "Airports vs. Airlines" metaphor, where airports are the open infrastructure that allows airlines to operate. How does CKAN fit into that analogy?

Joel: It’s a powerful metaphor - more so as it comes from Miller Abels - Deputy Director and Principal Technologist of the Gates Foundation where they used CKAN to build the Gates Data Exchange - their Research Data Management system, for which I was fortunate to have been involved in while I was at Datopian.

And Miller puts it best in his “Airport vs Airlines” metaphor:

Applying that metaphor to CKAN, it is exactly what Miller called for -

an open source platform built using open source, that uses open standards to help host and exchange disparate data assets coming from various systems - from exports from long-defunct mainframe applications to the latest proprietary cloud-based application - without “denaturing markets”, as they all emit data and metadata that needs to be interoperably exchanged in a secure, standards-based manner - with CKAN serving as a robust “cooperation space.”

And like Airports, no single vendor has a lock on CKAN technology. You can even spin an instance yourself if you’re willing to invest the time.

If you’d rather focus on populating and curating your data catalog, it’s also available from several providers around the world.

Trust, FAIR data, and verifiable infrastructure

Let’s turn to interoperability — often framed as a technical requirement, but in practice, it’s deeply social. Interoperability isn’t just about JSON schemas or API endpoints, right? It’s about systems talking, but also about people, institutions, and values aligning. What does CKAN’s role look like in that landscape?

Joel: You’re right, interoperability is deeply social.

Beyond communications and interoperability at the technical level, CKAN also has a social layer.

Already, through the hard work of the Open Knowledge Foundation and the CKAN stewards, it has gained the trust of the community as the largest open-source data management platform.

With the CKAN POSE project, we are co-creating more social scaffolding with the Ecosystem Catalog in collaboration with the POSE Steering Committee.

Apart from cataloging CKAN sites and extensions, we’ll also catalog data standards, corresponding metadata schemas, tools, and applications that work with the CKAN API.

We’ll also host conversations around all these entities with an integrated forum - where users can also have discussions in user-led affinity groups about all topics Data.

Digital sovereignty and open governance

Joel, the concept of "digital sovereignty" also came up in your updates. What are the key implications of this for open source and DPI on a global scale, and how should CKAN navigate this landscape?

Joel:

It’s all about Freedom.

Open Source is not just about Free Software - it’s also about Freedom to Use, Study, Modify and Distribute software. Beyond code, it is also a model for decentralized software development that encourages open collaboration.

For DPI, “Digital Sovereignty” is about Freedom - freedom from proprietary vendor lock-in; freedom for a nation to maintain control, autonomy, and determination over its systems, data, and technological infrastructure.

Open Knowledge is about Freedom - freedom to use Knowledge that is free to use, reuse, and redistribute without legal, social, or technological restrictions. CKAN is at the intersection of these Freedoms.

As an Open Data Infrastructure, CKAN should create more pathways for the ecosystem to use Open Source techniques to democratize turning Raw Open Data into Open Knowledge that is Useful, Usable & Used towards DPI’s goals.

Joel, in one of your reflections, you quoted: “If we believe in trust, we have to earn it.” What does trust look like in code, specifically?

Joel: As I mentioned earlier - Open Source is also a software development model - where we co-create software in the open.

And CKAN has been open source from the very beginning - 12,000+ tickets over the past two decades, built issue by issue, by hundreds of contributors around the world! And it's not just for the core platform - the vast majority of CKAN extensions built by the Community are open-source as well.

We’re earning Trust in the Data we’re managing in CKAN too! With DOIs, checksums, and several DCAT3 experiments already underway on several CKAN instances.

With close coordination with different organizations deploying DPI and through our work in POSE, I’m hopeful we can further enable and democratize Trusted Data with CKAN.

Product–Community Fit and building with people

Nora, your LinkedIn post mentioned learning about "sustaining and growing community around open source projects." What are the most actionable lessons you gained about fostering and supporting open-source communities, especially for public good initiatives?

Nora: I think there’s an opportunity to have a more focused discussion on this subject at the next UN gathering.

But throughout the meeting, I did hear comments on growing community that resonated with two themes that I’ve heard in the CKAN listening sessions.

First, there are many ways to contribute to open-source projects, and those contributions can be beyond code development and maintenance.

Second, I heard attendees recognize the importance of developing strong documentation and processes for onboarding and supporting new and existing members of an ecosystem and creating recognition structures that highlight contributions to the ecosystem.

Joel, you quoted Nabiha Syed: "It's not just Product Market Fit. It's also Product Community Fit." For CKAN, what does this mean in practice, and what concrete steps can we take to build this "Product-Community Fit" for long-term impact?

Joel: My first civictech startup - Ontodia, was created shortly after we won Mayor Bloomberg’s NYCBigApps challenge. As engineers, my co-founder Sami and I were not conversant with the startup scene. Fortunately, we were able to get Ontodia into a startup incubator where we got mentored by VCs and had regular meetings with other Founders.

The conversations were peppered with Three-Letter Acronyms (TLAs) - Monthly Active Users (MAU), Annual Recurring Revenue (ARR), Total Addressable Market (TAM), Customer Acquisition Cost (CAC), Minimum Viable Product (MVP), and of course Product Market Fit (PMF).

PMF is often considered to be the most critical TLA for startups. Because we’re often tutored, that only after achieving PMF can you start iterating and scaling, and start focusing on Growth Optimization (AKA Growth Hacking)

As we were the only civictech startup (we were even told by one VC not to mention civictech in our pitch deck), we adapted these TLAs as well.
But there was always a nagging feeling that the typical “startup” vocabulary was ill-fitting for civictech/govtech.

• “Move Fast and Break Things” often meant exploiting loopholes if not outright breaking the law to gain market share before regulators could respond
• Optimizing for Growth with little or no consideration of Consequences - unintended or otherwise.
• Consumers not Citizens, nor Communities

That’s why as soon as Nabiha said “Product-Community Fit” (PCF) - my ears perked up and I posted it on LinkedIn immediately.

Apparently, PCF is a new metric. Unlike PMF, which focuses on individual customer satisfaction, PCF emphasizes the network effects and social dynamics that emerge when users connect, collaborate, and create value for each other through the platform.

In the CKAN POSE project, we’re also exploring activating CKAN’s latent network effect potential with the Ecosystem Catalog. By using GitHub and LinkedIn social logins, we aim to organically connect users across the ecosystem, and build on CKAN’s built-in federation features to connect CKAN sites (at their option) to the Ecosystem Catalog.

Joel, the philosophy of "building with people, not just for them" resonated in your updates. How does this apply to creating open-source solutions, especially when working with different groups like governments or community organizations?

Joel: #BuildWithNotFor was always something that resonated with me when the concept was first advanced by Laurenellen McCann in 2014.

To me, it’s the next logical step for open source development - where we build with intent - not just with other developers, but with our end-users as well.

Too many times, governments and community organizations end up primarily as project managers.

There is no capacity-building, never mind knowledge transfer to the client - a critical step for them to be effective co-creators on their journey to be digitally sovereign and self-sufficient.

With our implementations, we often tell our clients that it’s in their “enlightened self-interest” to contribute back their improvements to the project.

We tell them: focus on maintaining and curating an authoritative, high-quality data catalog - and doing storytelling on top of the corpus.

By contributing your improvements upstream to CKAN, you actually benefit if the project merges your changes. It gets maintained by the project on your behalf, and the entire ecosystem benefits from your contribution.

We try to incentivize this by attributing the contribution to the client (with their consent).

Several improvements that are now core CKAN features started this way - Data Dictionaries (originally built for Boston), XLoader (commissioned while we were at OpenGov) and Datapusher+ (supported by Texas) are but a few examples from datHere. The CKAN stewards do a form of this as well.

What if we do this at scale using a formal process to iteratively co-create DPI capabilities into CKAN?

Sustainability, FAIR data, and CKAN’s future

You were both at UN Open Source Week representing the NSF-funded CKAN POSE project, focused on strengthening the CKAN ecosystem. How do the insights from the week directly inform your strategic approach and priorities for the POSE project moving forward? Are there specific areas where you now see a greater need for investment or a shift in focus?

Nora: While I listened to the panels, it was clear that there were common challenges on this point (and shared interests in seeing more commitment from the government sector in open source).

I recognized the variety of approaches in the room - and the opportunity to learn from one another. I can imagine it could be a useful strategy to invite representatives from other open-source ecosystems to come together to co-create a collection of strategies for onboarding new participation and sustaining existing participation.

Joel: The timing couldn’t be better for the CKAN POSE project. During our Phase I interviews in 2023 and 2024, we heard fragments of the DPI conversation.

The UN conference really put a spotlight on DPI just as we’re executing on the implementation phase of the POSE project.

Fortunately, POSE is all about building “Pathways to Enable Open Source Ecosystems” and a lot of the planned scaffolding for Phase II will be “co-tweaked” to account for DPI.

You both mentioned sustainability. It’s a critical challenge for open infrastructure. How do we make open infrastructure last? And from your time at UN Open Source Week, what concrete strategies or models did you encounter that could help ensure long-term viability and public impact?

Joel: Once again, the timing is perfect for the CKAN POSE project as we’re currently discussing Sustainability with the POSE Steering Committee - where OKFN, the CKAN Tech Team, and the CKAN Stewards are represented.

It was gratifying to see that there’s recognition from open source stakeholders, most importantly - the funding entities, that there’s a responsibility to support the digital commons (Miller Abel’s “Airports vs Airlines” metaphor comes to mind).

There are several sustainability models currently being discussed, including mechanisms that incentivize and reward good open-source citizenship to help address the endemic “Free Rider” problem common in open source.

With the global push toward FAIR data — Findable, Accessible, Interoperable, and Reusable — how can CKAN evolve to better support data that’s not just published, but truly usable and impactful within public infrastructure?

Joel: CKAN already has FAIR Data support with Adrià extending ckanext-dcat to support DCAT3 last year. And thanks to his close collaboration with Ian, it was done in such a way so that additional DCAT3 profiles can be fine-tuned and implemented using configurable, reusable, sharable ckanext-scheming metadata schemas.

At POSE, we’re coordinating with the Steering Committee to start cataloguing these schemas as well to promote and encourage re-use in the CKAN ecosystem across CKAN instances.
Not only will this make Interoperability easier, but it will also allow CKAN sites to share apps and tools they built using the same metadata schemas!

With the rise of AI, FAIR metadata is the best kind of context for AI initiatives. Facilitating the easy creation of high-resolution, high-quality FAIR metadata for each dataset and as an interrelated corpus - can make CKAN an integral part of an organization’s AI strategy.

What are the most significant opportunities for CKAN within the evolving global Digital Public Infrastructure movement over the next 12-18 months, and what is the "next frontier" for CKAN in this context?

Joel: The DPI movement reminds me of the early days of Open Data when there was so much energy and optimism. Admittedly, a lot of that Enthusiasm met with Reality, as publishing open data took care of the “supply”, but there was no easy way to track and coordinate the “demand” for open data.

The thing I love about DPI is that it addresses the entire spectrum - from Vision through Application - or what I’m calling the “DPI Intentional Innovation Cycle” - As Vilas puts it:

“From Values & Principles to Code.
Code to Governance. Code to Policy,
and Policy back to Innovation”

CKAN already provides a solid foundation for building DPI solutions. Over the next 12-18 months, doubling down on DPI can take CKAN from a “nice-to-have” to a “must-have” for public infrastructure.

Final Thoughts

Was there a specific moment, discussion, or presentation at UN Open Source Week that particularly resonated with you?

Joel: Vilas’s Keynote speech was transformational for me. I’ve listened to it so many times and even transcribed it. It’s my DPI manifesto.

What was the most surprising insight or challenge discussed at the event that you hadn't fully considered before?

Joel: Sami and I learned a lot of lessons the hard way since we started our Open Data journey. Going from winning NYCBigApps in 2011; which we parlayed to starting our first govtech startup - Ontodia; to being acquired by OpenGov in 2016; to deploying CKAN at scale across the US - we saw how govtech was built, marketed, sold, and procured - and found it unsatisfying - not just for us; for the direct customers - government; and ultimately, for the real end-user - government constituents.

We started datHere in 2020 and intentionally refrained from blindly using the Silicon Valley playbook.

It’s gratifying to see a lot of our organic observations being validated at UN Open Source Week.

We’re also very grateful that we had the luck and trust of our partners - the University of Pittsburgh and the CKAN core stakeholders - to have a third chance at co-creating a new way forward for CKAN as a Digital Public Infrastructure.

The open data movement has matured. Where is the “soul” of the work now? What questions should we be asking — not just as technologists, but as humans building shared digital futures?

Joel: Technology has made the world smaller. Globalization, Automation, and Financialization have lifted billions from destitution, but at the same time, have had Unintended Consequences.

With the emergence of AI, a lot of folks have been commiserating on a dystopic vision of Mass Joblessness - not just for Factory Workers, but Knowledge Workers too.

I’m hopeful that instead of Dystopia, AI will catalyze the required systemic change to build a Better, Equitable, Abundant Digital Future - with DPI initiatives playing a key role in regaining Trust, Citizenship, Sovereignty, and Global Stewardship.

What would you say to someone — a policymaker, a student, a local government worker — who wants to be part of this shift toward digital public infrastructure and open systems? Where can they begin?

Joel: Just jump in! You don’t need to be a developer. Data touches everyone.

You can contribute your perspective, your talent, your feature requests, your bug reports, your data stories. You can help clean the data, help with the documentation, and help others participate in the ecosystem.

At the end of the day, the product is a healthy ecosystem!

Finally, what is one main message or call to action you'd like to share with the CKAN community and our readers, based on your time at UN Open Source Week?

Joel: Come join us! Everyone can play a part in co-creating our Digital Future!