Idealism in Practice: How to Make Open Civic Tech Sustainable

Yuting Jiang September 22, 2026

Open-source civic tech creates public value, yet many teams who build it struggle to keep it alive. On 22 September 2026, we took that problem to a Metagov community call, together with Jean-Lou Fourquet of Lyfe Catalyst.

We came as a builder and a practitioner. Agora Citizen Network is an open-source deliberation platform that helps groups move from disagreement to a shared decision. Lyfe Catalyst designs deliberative processes for public institutions and companies. Here is what we shared, what we are betting on, and what we still have not solved.

Watch the full recording on YouTube

Two Journeys, One Meeting Point

Agora did not start as civic tech. It began as an anonymous forum at a school in France. Using zero-knowledge proofs, students could prove they belonged to the school without revealing who they were. That made it safe to talk about pay, workplace experience, integration, and dissatisfaction with school life.

It also showed us the limits of an ordinary forum. A thread on whether students regretted joining the school quickly turned tense. We considered using a language model to warn people when they sounded aggressive, then dropped the idea: participants would rightly feel policed. Discovering Polis and its use across the world pointed us to a better path, which is to map how people vote, reveal opinion groups, and identify statements that bridge the gaps. That shift, from social network to citizen network, became Agora.

Jean-Lou arrived from another direction. Through climate communication and his work with Tournesol, he concluded that more information was not the bottleneck. Recommendation systems decide what gets amplified, and they tend to reward emotion and polarization. What groups lack is coordination. That conviction led him to co-found Lyfe Catalyst.

We met in the middle: a builder looking for partners, not just clients, and a practitioner convinced every organization needs tools like this.

Finding Product-Market Fit Together

In an Agora conversation, participants react to short statements with agree, unsure, or disagree, and add their own. Facilitators provide a few seed statements so nobody faces a blank page. Agora then shows the opinion groups that best explain the conversation, with AI-generated summaries of what characterizes each one. The most useful output is often the set of statements that every group supports despite their differences: common ground a community can build on.

Real deployments taught us what came next.

As Jean-Lou and other facilitators used Agora, a question naturally emerged: after a group reaches consensus, what next?

Prioritization Mode is our answer: participants compare small sets of options, and Agora produces a shared ranking. Conversation Mode shows where a group agrees; Prioritization Mode shows what matters most now. See it in action in Can 30 People Agree on AI Rules in a Day?, a workshop we ran with Jean-Lou in which an AI startup co-wrote and prioritized its own AI charter.

Each deployment shaped the product, and each feature then served every facilitator, not only the client who asked for it.

Sell to Companies, Fund the Commons

Lyfe Catalyst is testing one answer to the funding gap. The deliberative tools and processes built for the public sector work in companies too, and companies can pay for them. Lyfe Catalyst sells facilitation and consulting to organizations, then directs 10% of eligible revenue back to the open-source tools it relies on. Eligible means after direct project expenses and before salaries. In its first year, the intended beneficiaries are Tournesol and Agora, with room for more.

It is not the single answer. It is one attempt to make value flow from organizations that can pay for democratic innovation to the communities building the underlying commons.

On our side, we draw a line between clients and partners. A client is interested in a tool. A partner believes in the team, brings real use cases, and helps the product grow. For an early-stage civic tech project, a few close partners are worth more than many vendor relationships. The market for deliberation keeps evolving, so staying close to facilitators is how we learn what groups need next.

Open, Interoperable, and Sustainable: Can We Pick Three?

Open source alone does not make an ecosystem. Separate deployments of the same tool can stay isolated, so communities working on similar issues never learn from each other. Our strategy rests on two things.

  • Partnerships. We work with facilitators who bring the process, and with builders whose tools complement ours, instead of building every feature in-house.
  • Interoperability. We proposed the Decentralized Deliberation Stack (DDS), a way for deliberation tools to interoperate at the protocol level. It is now a W3C Community Group, open to other approaches.

This comes with a real trade-off. Openness and interoperability lower development costs for builders, and they also lower switching costs for clients. A client can leave us more easily than they could leave a closed platform. We have not resolved that tension. Our bet is simple: trust holds people better than lock-in.

What the Community Added

After the presentation, breakout rooms took on two practical barriers: turning participation into institutional action, and making civic tech something people actually want to use.

On institutional buy-in, participants suggested:

  • Involve public officials early, through small and low-stakes demonstrations.
  • Find champions across party lines.
  • Work with implementers, professional associations, and the people who write model policy, not only elected representatives.
  • Start with participatory budgeting, because it is tied to a real allocation of money.

Decision-makers also need confidence that a process will be legitimate and manageable. In polarized or conflict-affected settings, trust, security, and accessibility can matter as much as the platform's features.

On adoption, Jean-Lou pointed to Agora's mobile experience and to the curiosity that AI-generated opinion-group summaries create. Our view is that civic tech cannot treat design as decoration. Participant-facing tools must be at least as appealing as commercial apps. Advanced analysis can stay with facilitators, while the participant experience stays simple, social, and engaging.

We Haven't Solved This

We ended the call with two open questions, and they remain open for us:

  1. Openness is a public good; capture is private. How do we stop the value of open tools being captured by whoever packages them best?
  2. How should value flow back to the people who build the commons?

Lyfe Catalyst's redistribution model, our partnership approach, and DDS are partial answers. None of them is enough alone. If you facilitate deliberative processes, fund democratic infrastructure, or build complementary tools, we want to hear how you approach these questions—and to find out whether we can answer them together.

Try Agora · Watch the call · Lyfe Catalyst