Daily Mandi Price Retrieval through Open Government APIs: A Pilot Framework for Digital Agricultural Advisory in India

Abstract

Aims: To document a lightweight pipeline for retrieving current-day wholesale mandi prices from the Government of India's data.gov.in AGMARKNET resource and to assess its suitability as a potential data layer for digital agricultural advisory. Study Design: Applied technical pilot using a single cross-sectional API retrieval. Place and Duration of Study: Department of Statistics, Mathematics and Computer Science (CIMCA), Sri Karan Narendra Agriculture University, Jobner, Rajasthan; retrieval and analysis were conducted on 19 August 2026. Methodology: A Python client queried the resource "Current Daily Price of Various Commodities from Various Markets (Mandi)" (resource ID: 9ef84268-d588-465a-a308-a864a43d0070). The API reported 13,181 nationwide records for the date and returned up to 1,000 records on the retrieved page. A manually delimited, non-probability subset of 104 records from the first page was retained. Records were checked for field completeness and ordered prices. For each market-commodity-variety-grade-date record, a custom recorded price-range ratio was calculated as (maximum price - minimum price)/modal price x 100. Results: The 104 retained records covered 13 states, 30 markets, and 53 commodities. The mean recorded price-range ratio was 22.7% (median 15.1%; standard deviation 24.3%), with values from 0.0% to 103.9%. Eighteen records (17.3%) had identical minimum, maximum, and modal prices. State and commodity-category summaries were descriptive only because the subset was neither exhaustive nor probabilistically sampled. Conclusion: The live API supported programmatic retrieval and elementary validation of current-day mandi prices. The pilot does not evaluate an artificial-intelligence system, farmer use, temporal volatility, causal effects, or national representativeness. An operational advisory application would require exhaustive or prespecified sampling, scheduled longitudinal retrieval, immutable data and code archiving, and user-centred validation.