Varsha Shah is an enterprise technical architect at Tata Consultancy Services working for Microsoft on financial reporting, compliance, and intelligent automation. Her research tackles a fundamental weakness in corporate oversight: payroll records, invoices, tax filings, and procurement transactions can each appear legitimate while collectively exposing fraud or regulatory violations.
Shah’s enterprise finance experience spans payroll compliance, employment-tax monitoring, procurement controls, and audit automation. In 2026, her financial-compliance research proposed connecting those traditionally siloed systems through cross-document financial intelligence.
A connected approach to enterprise risk
- Graph-based entity correlation: Connects employees, vendors, accounts, transactions, and regulatory filings to reveal discrepancies invisible within individual records.
- Adaptive probabilistic risk scoring: Combines anomaly strength, source reliability, historical behavior, and completed investigations to prioritize credible threats and improve subsequent decisions.
- Cross-jurisdictional normalization: Reconciles currencies, tax requirements, reporting periods, and classification schemes so financial activity can be evaluated within its proper regulatory context.
- Explainable compliance intelligence: Uses language models to interpret contracts, policies, and audit notes while keeping evidence, investigator oversight, privacy, and access controls central to consequential decisions.
Shah reported testing this architecture against approximately three million records spanning five years and four jurisdictions, achieving 91 percent precision, 87 percent recall, and an F1 score of 0.89. She also reported 76 percent fewer false positives and 40 percent less manual auditing. Her AI Engineer World’s Fair presentation emphasized integrating these methods with existing payroll, procurement, tax, and enterprise resource-planning systems.
Her AI-enabled supply-chain architecture repository extends that enterprise focus to operational problems and implementation blueprints. Throughout her work, Shah advances predictive financial governance: identifying connected risks early enough for human specialists to intervene.
