Pilot
Bring one Jira story. See reviewed test cases and a draft Playwright PR.
A pilot shows, on your own application and your own stories, whether AI Testing Probe saves your team time. Your analysts and engineers stay in the loop, every AI step waits for their approval, and you decide on measured results.
Before we start
Prerequisites: a page object library and test IDs. If they are missing, setting them up is the first pilot step.
Scope
One application, a few real stories
We keep the pilot small enough to finish and real enough to judge.
-
One application
A single web application your team already tests, in your environment.
-
A few real stories
Stories from your current backlog, not a prepared demo set.
-
Your people in the loop
Analysts approve test cases, engineers review pull requests. Nothing merges without them.
Duration and effort on your side: [Pilot duration and customer-side effort, to be confirmed by our commercial team]
How a pilot runs
Five steps from kickoff to decision
-
Prepare the basics
If your page object library or test IDs are missing, setting them up is the first pilot step.
-
Connect in your environment
We deploy next to your Jira and repository. You decide which data reaches the model.
-
Work the stories twice
The same story sets are handled with AI and without AI, by the same kind of people.
-
Measure and report
Your analysts time both paths. Results are reported monthly, gates included.
-
Decide
You decide on the measured results whether to continue, adjust or stop.
How we measure
The same story sets go through your usual process and through AI Testing Probe. Your analysts record the time spent on test case preparation and script development in both. We do not use numbers from other customers.
Acceptance gates
Gates we measure for the first time in the pilot
These are acceptance gates, not results we have already proven. The pilot is where they are measured for the first time.
- Pilot gate ≥ 70% of AI test cases accepted with minor edits (first measured in the pilot)
- Pilot gate ≥ 60% of scripts compile and run at first review (first measured in the pilot)
If a gate is missed: [What happens if a gate is missed, to be confirmed by our commercial team]
Targets
What we aim for, once a baseline exists
Targets, not results. They are confirmed only after your analysts have measured the baseline.
- Target 30–50% less test case preparation time (target, confirmed after baseline measurement)
- Target 20–25% less script development time (target, confirmed after baseline measurement)
Commercial model
A fixed-scope pilot, then a licence
The usual shape is a fixed-scope pilot followed by a project or platform licence. We do not publish prices; we agree on them after the fit call.
[Commercial model wording, to be approved by our commercial team]
Ownership
What you keep
-
The test code is yours
Generated tests are standard Playwright TypeScript in your repository.
-
The test cases are yours
Approved test cases are published to your Jira or Xray; the platform itself runs in your environment.
-
No lock-in
If you stop, the tests keep running in your CI without AI Testing Probe.
Plain Playwright, your repo
Migration
Do you migrate our Selenium suite?
Not with the product, today. AI Testing Probe generates new tests from your stories. Moving an existing Selenium suite to Playwright is on the product roadmap, and Antrek Labs offers it today as a service.
Roadmap
Who you work with
People and support behind the pilot
Team
Names and roles are published only with each person's written consent.
Team introductions follow during the first call.
-
Antrek Labs experience
AI Testing Probe is built by Antrek Labs, a test engineering company that also delivers web, desktop and mobile test automation as a service.
-
Reference call under NDA
On request, we arrange a reference conversation under a non-disclosure agreement.
-
Support and SLA
[Support and SLA summary, to be confirmed by our commercial team]
Bring one story.
In a 30-minute call we look at your stack and a few of your stories, and tell you honestly whether a pilot fits.