Data Reconciliation Script

You are a data analyst. Help me build a reconciliation between two data sources.

Source 1: {{system_name}}
{{key_field_amount_field_date_other_releva}}

Source 2: {{system_name_2}}
{{same_structure_from_second_source}}

Match on: {{key_field_invoice_number_transaction_id_}}

Produce a reconciliation that shows:
1) Matched items (exist in both sources with same amount)
2) Matched with differences (exist in both but amounts differ — show variance)
3) Source 1 only (exists in Source 1 but not Source 2)
4) Source 2 only (exists in Source 2 but not Source 1)
5) Summary statistics:
   - Total records and $ in each source
   - # and $ matched exactly
   - # and $ matched with differences
   - # and $ unmatched on each side

Format: Reconciliation report with sections for each category. Include variance analysis for matched-with-differences items.

Copied to clipboard.

Why this works

Cross-system reconciliation is manual and error-prone. AI structures the matching logic and produces a clean exception report.

Risks & review

Risks: Matching logic depends on data quality in both sources. Validate that the key fields are actually unique and consistent. Control: Finance reviews all unmatched items.