Certyn runs your test cases in a real browser, on a schedule or on every deploy, and files what breaks with a video and steps that reproduce it. Your cases can stay in TestRail, Zephyr, Xray or Azure Test Plans.
Sounds familiar
No migration and no rewrite. Certyn pulls the cases at run time, so your test management tool stays the source of truth.
Bring the tests you already have
Connect TestRail, Zephyr Scale, Xray or Azure Test Plans and pick the suites you want. Each suite becomes something you can run, schedule or trigger from CI. If you have no suite yet, Certyn explores the app and drafts the cases for you.
Say what runs and when
Group cases into a process, tag the critical ones, and set a schedule. Nightly on staging, every deploy, or both. Runs go in parallel up to your plan limit.
Certyn executes the steps in a real browser
It signs in with the environment test account and follows the case the way a person would. Results are pushed back into your tool as a run, one result per case, and repeated pushes never double-post.
Passed → Passed · Failed or Error → Failed
You review findings, not click-throughs
Failures arrive as issues with a video, the failing step and repro steps. A bug stays open until a person closes it, even if the next run passes, because a pass after a failure can be timing.
Tests are plain language. This is a complete one, ready to paste into a new test case.
Before you start
Go to the login page.
Sign in using LOGIN_USERNAME and LOGIN_PASSWORD
from the environment variables.
Add "Wireless Keyboard" to the cart.
Open the cart and enter the discount code SAVE20.
Verify the discount row appears and the total drops by 20%.
Verify no error banner is shown.Two rules make a case reliable: one workflow per case, and an explicit expected result. Write "verify an error banner appears and contains Invalid password", not "verify it works".
Every case carries its own history: how often it passes, how often it flips result with no product change, and how long it takes.
Each run keeps what it saw, so a failure is a report you can hand to a developer as it is.
Video of the whole session
Every click and every wait, so the argument about what happened is over.
Screenshot at the failing step
The page as it looked when the expected result did not appear.
Steps to reproduce
The path the run took, ready to paste into the ticket.
Pass rate and flakiness score
Certyn measures how often a case flips result without a product change.
Retest on demand
Re-run a single case against a fix and compare it with the failing run.
Results back in your tool
TestRail, Zephyr, Xray or Azure Test Plans get the run without you copying anything.
It replaces the click-through, not the judgment. Deciding what is worth testing, reading a half-broken flow and knowing it will confuse users, designing the edge cases, and calling whether a release ships: none of that is on the machine. What goes away is running the same 47 checks by hand every sprint.
No. Connect TestRail, Zephyr Scale, Xray or Azure Test Plans and Certyn pulls the current cases at run time. There is no import and no second copy to keep in sync. You can also keep cases in Certyn if you prefer one place.
Environment variables hold the test account and any seed data, and the case can reference them by name. For anything the UI cannot set up, seed it before the run the same way you do today, or let the case create what it needs as a first step.
Runs work from what is on the screen rather than from selectors, so a moved button or a renamed class is usually followed without a change. When the flow itself changed, the case fails and you update the wording, not a page object.
Flakiness is scored by how often a case flips result with no product change. Above 20 percent over at least 5 runs, a case becomes a quarantine candidate. Quarantined cases still run, but they count as neutral, so they cannot pass or fail the whole run.
Yes. The App Explorer walks the app without a script, reports what looks wrong, and proposes new cases for the gaps it finds. It is a second pair of eyes on the parts nobody wrote a case for.
Connect a test management tool
TestRail, Zephyr Scale, Xray and Azure Test Plans, with write-back.
Writing effective tests
The rules that make a natural-language case run the same way every time.
Scheduling tests
Nightly regression, per-deploy smoke, and what to run where.
Quality metrics
How pass rate, flaky rate and quarantine are calculated.
Start on the free tier, no card. Connect a suite or let Certyn draft your first cases.