India's district courts have 4.82 crore pending cases. That number surfaces briefly when an exceptionally old case concludes, or when an annual report lands. Then it disappears again into the background hum of a system everyone knows is overwhelmed and almost no one can write about specifically.
Shalaka built Justice Delayed out of that frustration. As a court and crime reporter for Hindustan Times Pune, she had ground-level intuition about the backlog but nothing she could quantify, localize, or turn into a story. Official data was too aggregated to be useful. A district total tells you nothing about which courts, which case types, or which hearings have been pending the longest.
Justice Delayed, developed with support from the Brown Institute for Media Innovation's Magic Grant, restructures district court data using individual case hearings as the smallest unit of analysis. That shift, from aggregated numbers to individual hearings, is what made exploratory visualization possible and useful.
This talk walks through what that looks like in practice: identifying courts with disproportionately old cases, spotting patterns in hearing schedules, and localizing pendency in ways that a ground reporter can actually act on. It also covers the gap between a large, messy dataset and a publishable insight, and how visualization helped bridge it.
The deeper argument is about method. Visualization here is not the end product of an investigation. It is the tool for figuring out where to look. If you work with large public datasets and spend more time lost in them than reporting from them, this talk is for you.