The tooling has never been better. Drag-and-drop builders, low-code platforms, LLMs, chart libraries, templates for every use case. Government departments, NGOs, and civic tech projects are all shipping public data dashboards now. On the surface, this looks like progress.
Aman thinks most of it is the wrong kind of progress.
The problem is not technical skill or bad intentions. It is inherited assumptions: ideas about what a dashboard should be, absorbed from enterprise BI tools and never questioned for public contexts. The Big Ass Number. The Donut. The Bars. The download button. The assumption that if the data is visible, it is accessible.
Most public dashboards get built, launched, maybe written about, and then sit there waiting for the next person to come along and build the same thing again. They look helpful. Put them through their paces and the complexity turns out to be shallow.
This talk is about the gap between the dashboards we keep building and the public data infrastructure we actually need. Drawing on experiments at Diagram Chasing and a survey of interfaces from national portals to scrappy civic tech projects, Aman walks through a critique of the current landscape, the design assumptions worth questioning specifically for public contexts, and the technical possibilities that are not being used nearly enough.
The FOSS ecosystem is better than ever. DuckDB WASM, PMTiles, MapLibre, parquets in the browser, webR. What would it look like to build public dashboards that cost nothing to host and outperform enterprise tools? What are the small but consequential decisions that shape whether a data interface actually serves the people it is built for?