Skip to content
Back to use cases

Software vendors

Agent-driven regression testing

Acceptance test scenarios written in Excel by business teams become browser tests replayed automatically for each user profile. Every failure comes with its justification and screenshots.

Illustrative mock-up of the tool : acceptance · replayed tests
Illustrative mock-up: the names, values and documents shown are fictitious.

Context

An AI platform that changes every week, tested manually across several user profiles.

The problem

Lengthy manual acceptance testing and regressions discovered late.

What we built

A testing platform: an agent explores the application and writes deterministic Playwright scripts once, then the tests run without AI in the loop, in parallel batches, with a pass/fail report and evidence.

Steps

  1. 01Import of the business scenario catalogue
  2. 02Script generation by an agent through browser exploration
  3. 03Batch execution and reports
  4. 04Triage campaign for flaky tests

Hosting and models

Client's Azure cloud (Container Apps, Azure SQL); scripts generated by Claude Sonnet; execution without a model.

Services involved

Forward Deployed Engineers · Training & maintenance

We name a client only with their written agreement. The budgets and detailed results of our engagements remain confidential.

Tell us about your use case.

An engineer replies within one business day with the scope of a first use case and a schedule.