02 · Pegasystems · 2021–2023

400 Hours Back

QA teams at Pega were losing a week of every release to manual regression checks. The fix wasn't a better testing tool. It was no separate tool at all.

400hours back, every release

$3.1M+saved per year

40%faster QA cycles

Role: Senior UX Designer. End-to-end design

Test Studio inside Pega Agile Studio: an epic showing 37 tests with pass, fail, and blocked counts, plus the test plans behind them — The finished thing: test coverage living on the epic itself. No attachments, no spreadsheet, no second login.

The problem

Development happened in Agile Studio. Testing happened everywhere else: test plans as attached documents, results in spreadsheets, coverage reconstructed by hand at the end of every sprint. By the time anyone could say "this epic isn't tested," the release was already late. Each cycle burned roughly 400 hours of manual regression work.

The old workflow: a QA epic with 43 stories, test plans attached as documents, and results tracked outside the tool — Before: test plans as attachments, results somewhere else, and a QA lead doing archaeology every sprint.

The decision

The obvious fix was a dedicated testing product. That was also the wrong fix: one more login, one more silo, one more place for the truth to drift. The decision was to make testing part of the work itself, with plans, executions, and coverage living inside the same epics and stories developers already worked in. If a developer never has to leave the epic to know whether it's safe to ship, the tool disappears.

The decision, drawn: kill the detour, put the tests on the rail. Coverage becomes something the epic has, not something QA reconstructs.

How I got there

I shadowed QA leads through a full release cycle and mapped every handoff: where testing information was created, where it went stale, and who paid for the gap. The journey map made the argument better than any deck: five stages, and the pain concentrated exactly where testing left the development tool.

Journey map of a release cycle across planning, execution, reflection, resolution, and completion — A release cycle, mapped. The pain lives in the gaps between tools.

Before-and-after journey: the old path scattered across tools, the new path smooth with 400 hours recovered — Before and after. The 400 hours didn't vanish. They moved to testing that actually needed a human.

The design, in motion

A short walkthrough of the shipped experience: creating a test plan against a sprint, executing it, and reading coverage without leaving the epic.

Two minutes of the real product. Click to play. It won't load until you do.

What happened

400 hours returned per release cycle, about $3.1M a year across teams, and QA moved 40% faster. The deeper change: testing stopped being a phase at the end and became a signal available all the way through. The consistency standards from this work were adopted by five product teams across the division.