Cities, software, and the people in between

A city runs on systems that were never designed to talk to each other: transit, housing, permits, utilities, public space. Software promised to unify them, and in places it has — but the interesting lessons come from where digitization met reality. The counter where a resident waits is the ultimate user test, and it is unforgiving.

Technology succeeds in urban contexts when it respects the workflows of the people who operate the city daily. A planning tool nobody at the counter uses is not a tool. A dashboard that ignores how departments actually coordinate is decoration. The clerk who processes fifty permits a day knows more about the real workflow than any requirements document, and the best projects we have seen start by shadowing that clerk for a week.

The emerging discipline looks less like app development and more like institutional design: mapping who decides what, where information stalls, and which small improvements compound across thousands of daily interactions. A single shared reference — one address record, one application ID, one status that every department trusts — routinely outperforms a grand platform rewrite.

Start narrow and visible. The intersection residents complain about most, the permit type with the longest queue, the bus route where bunches cascade. Fix one corridor end to end, publish the before-and-after numbers, and let the credibility fund the next corridor. Cities learn like products do: slice by slice, in public.

Data helps most when it shortens a specific decision. A depot manager deciding how many buses to hold, a planner deciding whether a junction needs a signal retime, a housing officer triaging inspections — each needs a faster signal, a clearer tradeoff, and a shared fact the other departments will accept. Dashboards that serve those three moments get opened daily. The rest get opened for inaugurations.

Equity has to be designed in, not audited afterwards. Digital-first services quietly exclude the residents who need the city most unless every online step has an assisted equivalent at the counter, in the local language, with the same service guarantee. The cities getting this right measure completion rates by ward and income, not just aggregate uptime.

None of this replaces the judgment of people who run cities. It gives that judgment better material — faster signals, clearer tradeoffs, and shared facts across departments that previously negotiated from different spreadsheets. Software becomes the meeting room where the city thinks together.

Software will not replace the judgment of people who run cities. It can, however, compound it. Done well, that partnership — institutional knowledge plus machine-readable fact — is what modern governance infrastructure actually means.