Context/Background

A financial management system processes transaction data from multiple sources (payment gateways, banking APIs, accounting software) to generate quarterly financial reports for regulatory compliance and executive decision-making. Data quality issues in source systems have previously led to reporting errors that required manual correction and restatement of financial results.

Source

Financial transaction data is collected from various internal and external systems and must be validated before being used in regulatory reporting.

Metric/Acceptance Criteria

The financial data processing system must ensure:

• 99.99% accuracy of financial transaction data (maximum error rate of 1 in 10,000 transactions)
• 100% completeness of mandatory fields for all financial records
• Automated validation checks must flag at least 98% of data anomalies before data enters the reporting pipeline
• All data quality issues must be logged with severity levels and resolved within timeframes based on severity:
  • Critical: 1 hour
  • High: 4 hours
  • Medium: 24 hours
  • Low: 72 hours

This example was created with help from Anthropic-Claude-Sonnet-3.7 with the following prompt:

I want to add an example requirement (under _requirements) for data-quality. 
It shall follow the pattern "Context/Background", "Source", "Metric/Acceptance Criteria" and contain precise metrics 
for acceptance criteria.