India's official statistical system produces an enormous volume of data across hundreds of ministries and departments. Most of it has been difficult to find, harder to compare, and nearly impossible to use without specialist knowledge.
eSankhyiki is MoSPI's attempt to change that. A unified digital platform bringing together over 700 datasets and 136 million records into a single accessible interface. It is built on a core premise: visualization is not a presentation layer added at the end. It is part of the data lifecycle itself, from discovery and validation through to interpretation and policy action.
This talk walks through the design, development, and deployment of the portal. Rohit covers how user research shaped the visualization choices, how thematic dashboards and interactive filtering tools help users move from broad questions to specific insights, and how metadata-driven interfaces make complex socio-economic data navigable for audiences ranging from policymakers and researchers to journalists and citizens.
You will hear about the constraints that shape visualization in government and public sector environments: scale, data sensitivity, accessibility standards, and the need for long-term sustainability. And you will hear how those constraints were balanced against clarity, transparency, and trust, values that the team treated as non-negotiable.
The talk also touches on AI-enabled semantic search and how it complements visualization by improving data discovery before a user even reaches a chart.
For anyone building visualization systems at scale, or trying to make public data genuinely useful to the people it describes, this is a rare inside account of what that work actually involves.