SAP Test Automation With AI Agents: How Glasslake Runs SAP Tests on Browser Use
Browser Use published a case study on how Glasslake runs SAP test steps with Browser Use Cloud agents. This is a customer walkthrough of an existing setup, not a new feature.
Glasslake is an agentic AI platform for SAP transformation projects. Its own agent prepares the test scenarios. A Browser Use agent then does the browser work in SAP Fiori: it signs in to the chosen SAP environment with a test user, finds the app named in the test step from the launchpad, fills in fields and completes the transaction, and checks what SAP shows back. Unlike a recorded script that points at fixed fields, the agent works from the written test step and reads the screen to decide each action.
Each test starts a Browser Use Cloud agent run in a managed cloud browser. Glasslake gives the agent task instructions for SAP, such as how to open an app from the launchpad and when a step counts as done. It embeds the live browser view in its own product, so a consultant can watch the run as it happens. Glasslake also attaches its own client to the same browser over the Chrome DevTools Protocol and captures screenshots as test evidence.
Long end-to-end SAP processes can take more than a hundred actions. Follow-up runs can continue in the same session, keeping the conversation and, while the browser stays active, its open pages and login.