96% Faster Data Reconciliation - Zero-Guesswork Validation for one of the Top 5 Canadian Bank's Global Derivatives Business

One of the Top 5 Canadian Banks

• The bank's Global Derivatives/FICC business generated a massive volume of trade data every day from different source trading systems, all loaded into a central database. Validating each record by hand made QA effort during SIT and UAT significantly high, and the exercise only grew harder as data arrived in multiple formats - Excel, JSON, and delimited DAT files - changed in real time, and grew too large to reconcile manually with any confidence. The bank needed a way to compare source and target data completely, not on a sample, and to see clearly where and why records differed.

• Use the DQV product to compare two DAT files, each approximately 1.5 GB in size, containing 800K records with 151 fields per record, end-to-end in 153 seconds, with no manual extraction or scripting required.
• Implement built-in aggregation across multiple key fields (10+) and calculate/derived-field expressions evaluated inline during comparison, completing the comparison in 243 seconds even with the additional processing.
• Validate wide Excel extracts at both ends of the spectrum -174K records × 72 columns in 151 seconds and 27K records × 216 columns in 51 seconds-covering both high-volume and high-column-count files
• Parse and reconciled nested JSON documents (3.2 MB, 68 fields, 6.5K records) in as little as 11 seconds.
• Generate detailed Excel reports for every comparison - DAT, Excel, or JSON - including mismatch statistics and a full record-level mismatch breakdown, eliminating the need for separate tooling for different formats.
• Add custom computed fields to the comparison results to calculate the deviation percentage and identify whether the deviation exceeded a defined alert threshold.
• Execute comparisons across the entire data set for every run, regardless of file type or size, with no sampling.