AI-Facilitated Sensemaking as Civic Infrastructure with Artem Zhiganov
Facilitation is the bottleneck to democracy
Jürgen Habermas — who had just passed when Artem joined us in March — spent his career arguing that genuine public deliberation isn’t a luxury. It’s public infrastructure, as legitimate as roads, plumbing, or libraries. Artem Zhiganov took that premise and built a product from it.
Harmonica is an AI-facilitated sensemaking tool that targets the most overlooked bottleneck in democratic practice: the skilled, expensive, irreplaceable work of a good facilitator. Surveys and forums scale, but they surface no depth. Interviews and citizen assemblies have depth, but they don’t scale. Harmonica’s thesis is that facilitation itself can be made composable — modular, deployable, AI-assisted — without losing the qualities that make it work.
In our session, Artem did something brave: he spun up a live Harmonica session on the spot, with the audience co-designing the facilitation prompt, deployed it to a Telegram bot, and had the whole thing running in under five minutes — all via a command-line interface and MCP server, with Opus as an intelligent co-designer. We then explored the broader ecosystem: the Open Facilitation Library (with MetaGov and Lisa Schirch), GovAc for Web3 governance, the Swiss public AI model Apertus, and the DDS.xyz Decentralized Deliberation Standard as potential connective tissue.
The critical question — raised by Tim in Q&A — is one worth sitting with: how transparent should the prompts in AI-facilitated civic processes be? Who audits the supply chain of the facilitation itself?
Artem Zhiganov is the founder of Harmonica. This session was recorded at the OpenCivics Bi-Weekly Network Assembly on March 17, 2026.