Output example
Before vs after is not enough. We compare execution.
Instead of visual comparison, schema validation and field mapping, we produce evidence that runtime behaviour still holds.
Execution trace comparison
Side-by-side alignment of source runtime stages and migrated runtime stages.
Step-level evaluation alignment
Formulas, groups, filters and dependencies are traced at execution-step granularity.
Runtime equivalence scoring
Calculation parity and semantic drift are measured instead of assumed.