Cross-browser automation for a store analytics dashboard

The client is a UK operator whose dashboard gives fulfilment teams a live view of orders, inventory, and reporting across a large store network. Those screens drive daily operational decisions, so a rendering or data fault carries real cost.

The dashboard had to look and behave correctly across a wide spread of browsers and devices, but nobody could verify that consistently by hand.

Manual checks could not cover the browser and operating-system matrix the dashboard's users actually ran, so layout and data faults slipped through on the less common configurations. Bugs were reported by users after release rather than caught before it.

The reporting and inventory views depended on API responses that were never validated end to end, so a backend change could silently break a chart. The UI and the data behind it were tested in isolation, if at all.

The operator needed dependable coverage across both the browser grid and the API layer, running before each release rather than after. Consistency across configurations was the requirement.

We built E2E automation in Cypress and Playwright over orders, inventory, and reporting, running across a 40-plus browser and OS matrix so regressions surfaced on every supported configuration. Covering the real grid is what stopped configuration-specific faults reaching users.

We paired the UI suites with Postman API checks, validating the responses behind each dashboard so a backend change could not silently break a chart. Pushing coverage down to the API layer kept the suite fast and broad at once.

Everything ran in GitHub Actions ahead of each release, gating the pipeline and reporting failures back with the offending configuration. Releases went out only once the grid was green.