Skip to content
Adarsh Dwivedi.AI & Product Engineer
02Work

Floodcast Gurugram

Route-Level Urban Flood Risk

Floodcast Gurugram running: Route-Level Urban Flood Risk.

Captured from the live deployment, not a mockup

01 · The problem

Answered the one question a commuter actually asks, will my route flood and when, instead of a city-wide alert they still have to interpret themselves.

02 · How it works

  • 01

    Made data confidence visible rather than hidden, with 39 of 73 flood points backed by a named source and a confidence value on every row.

  • 02

    Built the register so citizen reports feed back into it, turning users into the data source rather than only its audience.

  • 03

    Scored whether a specific route through Gurugram will flood and when, matching live rainfall against 73 researched flood points rather than issuing a city-wide warning nobody can act on.

03 · What it cost, and what it returned

Tracked provenance as a first-class field: 39 of 73 points carry a named source and every row has a valid confidence value, so a placeholder can never render as fact.