Issue
The client is an IPP on a major ETRM overhaul with need to convert complex price data between old and new schemas.
A price-data conversion program had been built but never tested:
- Differences between the old and new schemas had been spotted, but their full extent was unknown.
- Unvalidated conversions put the integrity of downstream pricing at risk.
- The scope of the differences had to be determined and an accurate solution delivered.
Solution
Our consultant reconstructed and validated the untested conversion methodically:
- Used the risk engine to produce testable data in the old and new systems.
- Reconciled the results and analyzed differences, grouping them into error classes.
- Built a conversion matrix that made each discrepancy easy to see and resolve.
Result
The price-conversion application was corrected and fully tested - and completed much sooner than the client expected - with the full scope of differences now known rather than assumed. In addition, the new price curve modeling resulted in much better reporting, analytics and P&L explain.
“Your consultant is brilliant!”
IT Manager
IPP
Key Takeaways
- A conversion you cannot reconcile is a conversion you cannot trust. Generating testable data in both the old and new schemas, then grouping the differences into error classes, turned an unverified price-conversion program into one that was provably correct.
- Structured reconciliation beats spot-checking. Building a conversion matrix from classified differences is a repeatable method for validating any schema or system migration where the full extent of discrepancies is unknown at the outset.
Ready to do the same?
Migrating price data between schemas and unsure what is breaking? MidDel's structured reconciliation surfaces and classifies every difference before go-live. Reach out.
