InterTwinkles

The bet

InterTwinkles bet that the software failures of online group decision-making were not really about scale, they were about the specific, structured needs of small groups trying to reach consensus, a co-op board, a housing collective, an activist working group, and that purpose-built tools for that narrow case would beat generic forums or voting software. It began as MIT Media Lab graduate student Charlie DeTar's dissertation project, developed in participatory design sessions with real Boston-area cooperative housing groups rather than built first and tested later.

Peak

The project matured into a genuine open-source toolkit, presented publicly through MIT's Center for Civic Media, and DeTar defended it as a full dissertation in 2013, a level of documented, peer-scrutinized design thinking few platforms in this catalogue ever receive. It was used by the actual cooperative and consensus-based groups it was designed alongside, a smaller footprint than most platforms here but a more genuine fit between tool and user than most achieve.

What worked

Building the tool through direct participatory design with the groups who would use it, rather than guessing at their needs from outside, produced software that reportedly fit how consensus-based groups actually make decisions, a real and undervalued discipline. The open-source release also meant the ideas and code outlived any single hosted instance of the tool.

What failed — the mechanism

InterTwinkles ran its authentication through Mozilla's Persona identity system, a reasonable choice at the time from a respected source, and one entirely outside the project's own control. When Mozilla later wound Persona down, the hosted version of InterTwinkles lost its login layer, and with no dedicated funding or team positioned to re-engineer around the change, the run service stopped rather than migrated. The lesson is not that InterTwinkles failed in the way the other cases here failed; it is that a small academic project has no slack to absorb a breaking change imposed by someone else's infrastructure decision, made for reasons that have nothing to do with the project at all.

What Agora takes from it

InterTwinkles is a reminder that platform risk is not only about funding models and users, it is also about which pieces of your own stack you do not control. A dependency on someone else's infrastructure is a dependency on someone else's roadmap, and a project with no resourcing to respond when that roadmap changes is exposed regardless of how good its actual design work is. Agora's answer is mundane but real: build and maintain the identity and participation layer as core, funded infrastructure, not as a convenience borrowed from someone else's discontinued service.