Automated Dataset Reconciliation
Problem
Organizations managing complex supply chains and GST compliance face significant challenges in manually reconciling multiple large datasets containing thousands of transactions. Specifically, businesses must reconcile three critical datasets: their internal purchase ledgers, government-provided GSTR-2B statements (auto-drafted Input Tax Credit eligibility), and Tax Deducted at Source (TDS) entries. This reconciliation process is critical for accurate Input Tax Credit (ITC) claims, tax compliance, and fraud prevention Automatically match and reconcile large financial datasets—for example, purchase ledgers with GSTR-2B reports or TDS entries—to detect and highlight discrepancies. Traditional reconciliation is time-consuming and error-prone because of varying invoice formats, missing records, partial matches, or rounding issues. Automatically align corresponding entries from both datasets using fuzzy matching (for invoice numbers, vendor names, or GSTINs). Identify differences in taxable value, tax amounts (CGST, SGST, IGST), invoice dates, and credit availability. Summarize mismatches by type (e.g., missing invoices, ITC mismatch, duplicate entries). Provide a reconciliation summary report and a detailed match-level explanation.
Prompt Input
Purchase ledgers, GSTR-2B data or TDS entries two or more structured datasets in formats such as Excel, CSV, JSON, or database tables: Purchase Ledger Dataset: Fields: Vendor Name, GSTIN, Invoice Number, Invoice Date, Taxable Value, CGST, SGST, IGST, Total Invoice Value, ITC Claimed. GSTR-2B Dataset (or TDS dataset): Fields: Supplier Name, GSTIN, Invoice Number, Invoice Date, Taxable Value, CGST, SGST, IGST, Total Tax Amount, ITC Eligible, TDS Deducted (if applicable). Optional fields: HSN/SAC Code, Place of Supply, Type of Supply. Additional parameters the AI may accept: Matching tolerance (percentage difference allowed in amount/date). Key matching fields (e.g., GSTIN + Invoice Number). AI confidence threshold for fuzzy matches.
Prompt Output
Identifying discrepancies in amounts, invoice numbers, and ITC availability. Reconciled Output Dataset with match results for every record, containing: Matched/Unmatched status. Discrepancy type (missing invoice, value mismatch, ITC mismatch, duplicate entry, date variance). Difference in amount, tax, or ITC claimed. Confidence score for fuzzy matches. Suggested corrective action (e.g., “verify vendor GSTIN” or “adjust ITC claim next month”). Summary Report: Total entries matched, partially matched, or missing. ITC differences in monetary terms. Vendor-wise or category-wise discrepancy summary. Export options: Excel/CSV/Database update.