Bizbrains automates UI testing for continuous delivery

Company Overview

• Company: Bizbrains.
• Industry: IT services / Software development.
• Business Type: Custom software development consultancy.
• Size: 50–100 employees.
• Geolocation: Sweden.

Key Use Cases

• Automated UI testing for continuous delivery.
• AI-assisted test generation (via Confluence Docs).
• Custom CI/CD integration (Azure DevOps via API).
• Complex data dependency setup & verification.
• Custom office dashboard reporting.

The Challenge: A UI testing bottleneck blocking continuous delivery

Bizbrains, a custom software development consultancy based in Sweden, set an ambitious goal to shift toward a continuous delivery model with automated rollouts every week or bi-weekly. While their engineering team had a strong foundation of unit and integration tests, UI testing remained heavily reliant on a manual, time-consuming process.

To achieve their CI/CD goals, Bizbrains needed to drastically reduce the time spent on UI testing, aiming to lower it to just 10% of their current manual effort. Furthermore, past attempts to automate these flows with other testing tools failed because the scripts were too brittle and frequently broke whenever the UI changed. They needed a resilient AI automation tool that could keep up with their rapid deployment pace and ensure their customers received a reliable product.

The Solution: Confluence-driven AI and Azure DevOps integration

Bizbrains ran a parallel proof-of-concept, pitting their normal manual UI testing process directly against QA.tech's AI agents during a major release. The development team quickly integrated QA.tech into their specific Microsoft-heavy workflow:

Test generation from confluence:

‍Instead of writing new scripts from scratch, the team imported their existing manual test case descriptions directly from their Confluence user guides into QA.tech's AI chat, allowing the agent to automatically generate modular test steps.

Managing complex entity dependencies:

‍Because the Bizbrains system requires extensive initial data setup – such as intricate document configurations requiring specific formats, variants, and partner routing – the team broke their tests down into modular, dependent test plans. This prevented massive end-to-end tests from becoming flaky and allowed the AI to reliably navigate their complex object hierarchies.

‍Azure DevOps & custom dashboards:

‍Bizbrains operates on Azure DevOps rather than GitHub. To shift testing left, their project lead, utilized QA.tech's API to build a custom Azure pipeline. This pipeline automatically deletes old test environments, provisions new ones with the latest product version, triggers the QA.tech test plans, and beams the real-time success or failure results directly to a custom big screen in their office for morning reviews.

The Results: UI resilience proven for 25,000 annual test runs

By adopting QA.tech, Bizbrains successfully modernized their UI testing layer, discovering that shorter, specific AI test cases performed exceptionally well for their purposes. Most importantly, the QA.tech agents proved highly capable of handling the visual fluidity that previously broke their older tools.

As Mikael Flensborg, Bizbrains' CTO noted:
“We’ve seen significantly stronger performance when handling UI changes – areas where other tools consistently failed.”

With the added ability to easily test across multiple browsers and screen resolutions automatically, Bizbrains is now preparing to build out a full suite of around 1,000 test cases. Running these weekly or bi-weekly will result in 25,000 to 30,000 automated test runs a year – ensuring high functional coverage and giving them the confidence to deploy continuously.

AI testing Highlights

Before: UI testing was highly manual and too slow to support their goal of continuous weekly or bi-weekly rollouts. Previous attempts at automation failed because older tools broke whenever the UI changed.

After: AI agents handle complex, multi-step UI testing, easily adapting to visual changes. Tests run automatically via a custom Azure DevOps pipeline and display real-time results on the office dashboard.