Error-Free by Design: AI‑Enabled TDS Compliance at EY India
DOI:
https://doi.org/10.5281/zenodo.17586932Keywords:
TDS compliance, PAN and challan validation, Section classification, Automation and dashboards, Exception governanceAbstract
EY India’s Global Compliance and Reporting Direct Tax team led a TDS engagement for a mid sized multinational processing over 2 500 vendor payments per quarter roughly 84 percent subject to withholding A manual pre check showed a six percent error rate PAN mismatches challan inaccuracies section misclassification and duplicates threatening Form 26Q timeliness and vendor trust The decision is how to stage a ninety day AI enabled redesign that automates validation and reconciliation through RPU OLTAS integration visualises anomalies via dashboards and embeds exception governance without slowing throughput The exhibit pack includes baseline error rates section wise distributions detection time and escalation ageing Options include a minimal validation stack a TDS Exception Probability Score control as checklist and targeted client training Learning objectives are to diagnose preventable errors quantify payoffs from automation design governance for judgment calls and plan a sequenced rollout The unresolved tension weighs model and change.
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Academic Research Publishers

This work is licensed under a Creative Commons Attribution 4.0 International License.