What Cettara is
One place for manual and automated testing, where every automated verdict shows its reasoning.
Cettara holds your test cases — the ones a person clicks through, and the ones a browser runs — in one place, and keeps the evidence for both.
The two kinds of case
A manual case is a list of steps and what should happen at each one. A person follows it and records what they saw.
An automated case is a list of steps a real browser carries out. Instead of writing code to assert the result, you write what should be true in plain language — "a draft invoice appears, awaiting approval" — and an AI judge rules on it against what is on screen.
A project holds one kind or the other. That keeps a manual pack readable and an automated suite runnable, and it is the first choice you make when you create one.
What you get back
Every automated run keeps video, a full execution trace and the screenshot each checkpoint judged. Every verdict comes with reasoning and the evidence it relied on, so you can check the work rather than trust it.
When the judge is wrong, you say so. Your correction is recorded, counted into an agreement figure you can see, and turned into a note the judge reads the next time it judges that same checkpoint.
How it is organised
- An organization is your company. It holds people, billing and the audit log.
- A project holds cases of one kind, plus the pieces they need.
- A run is one execution: of one case, of a flow, or of a milestone's worth of work.
- Evidence is what a run leaves behind, kept for as long as your plan allows.
Where to start
New to Cettara, testing by hand today? Start with Manual testing. Automating a flow? Start with Automation. Connecting an AI assistant? That is For developers.