Skyvern

Skyvern is the AI agent for resilient browser automation: it uses LLMs and Computer Vision to execute complex web workflows via a simple API, bypassing fragile XPath scripts.

Skyvern delivers robust, AI-native browser automation. Unlike traditional tools (Selenium, Playwright) that break on minor UI changes, Skyvern leverages Large Language Models and Computer Vision to ‘see’ and interact with websites like a human (no brittle XPath selectors). This open-source platform handles complex, multi-step workflows: think automating procurement pipelines, fetching invoices from hundreds of vendors, or completing dynamic government forms. You define the task via a simple API endpoint; the agent reasons, adapts to layout changes, and executes, providing a scalable, reliable solution for enterprise RPA.

https://www.skyvern.com/