Product
Start small, get better release by release.
You begin with fast baseline coverage, then deepen the check with rules. The same flow becomes more meaningful step by step, with no new rollout.
The flow
First secure, then verify
Three steps are enough to start: baseline, comparison, deviations. When you want a sharper check, the rule-based steps slot in at exactly the right places.
-
Verified baseline meets new actuals
You take the verified baseline from the last run and put this run’s actual results next to it. Result data lives in SQACE storage and serves as the baseline again next time.
Screenshot followsBaseline and current actual result -
Rule-based
Rule editor and a computed expected result
Instead of only checking against the last run, you define business rules in the visual rule editor, without SQL. From them SQACE generates the expected result as a second, independent calculation. That catches errors that were always wrong and never show up in any baseline.
Screenshot followsRule editor with the generated expected result -
Automated comparison
SQACE compares actual against baseline field by field and reports what changed that should not have. You set tolerances and key fields from the business side.
Screenshot followsField-level expected-vs-actual comparison -
Rule-based
Result breakdown by rule usage
The comparison shows the rule behind each expected value and breaks the result down by rule usage, including coverage. You see which fields are checked rule-based and where coverage is still growing.
Screenshot followsBreakdown by rule usage with coverage -
Classify deviations and reuse them
Automatic classification groups deviations by standard patterns and by your patterns from earlier runs. You assess only the rest, create findings and hand them to test management. The classification applies automatically on the next run.
Screenshot followsDeviations classified, drill-down to field and rule level
Product in motion
SQACE in two minutes
A short product animation of how business and IT use SQACE to check and compare data processing together.