The project developed a market monitoring tool for Unified Payments Interface (UPI) providers that leverages publicly available social media and application review data to surface, in near real time, the hurdles customers face. While UPI has rapidly become India’s dominant payment infrastructure, traditional complaint channels capture only a small fraction of user experiences, even though users frequently encounter onboarding challenges, failed transactions, and delays in grievance resolution.
To address this visibility gap, data was collected from Twitter and the Google Play Store for major UPI providers that together account for over 90 percent of transaction volumes. The data were modelled using natural language processing methods to identify an exhaustive set of customer hurdles, calculate their prevalence rates, and generate indicators of provider responsiveness and temporal trends. The tool, Dvara PARSE, is currently running, and it captures grievances daily and categorises them into themes like onboarding hurdles, inadequate grievance redress, unauthorised charges, etc. Further, the tool identifies overlaps across these grievance categories and quantifies patterns of customer sentiment and periodic traffic related to providers. The overall accuracy of the tool stands at 94 percent.
By automating the identification and measurement of customer-facing issues, the monitoring tool yields actionable insights for providers and regulators, enhancing the detection of emerging problems in the UPI ecosystem.
