Sauce Labs updated its AURA platform to let enterprises test software using open source, open-weight, or proprietary LLMs without changing release workflows.
Sauce Labs has announced the general availability of Bring-Your-Own-Model (BYOM) capabilities for enterprise customers using AURA, its agentic AI software quality assurance and release assurance platform. AURA now supports open-source, open-weight, and proprietary large language models (LLMs) as its underlying model layer.
The company decoupled the model layer from AURA’s release assurance intelligence, agentic workflows, and execution infrastructure, allowing enterprise customers to switch or mix and match LLMs without changing their underlying verification platform or CI/CD pipelines.
Sauce Labs cites industry data showing developers are generating 741% more code due to AI coding assistants, while software release velocity has risen by under 20%. Sauce Labs reports that enterprises using AURA achieve a 90%+ reduction in production incidents, 47% faster release cycles, and 38% of engineering capacity reclaimed.
The company takes a direct stance that enterprise context belongs to the customer, connecting directly to models and systems already tuned and governed by the customer. AURA leverages execution data drawn from over 8.7 billion test executions across more than 10,000 devices to verify software behaviour. The system captures production errors and feeds them back into business intent models to refine future test cycles automatically.
Sauce Labs was created by the original founders of the open-source Selenium and Appium testing frameworks. Sauce Labs supports over 300,000 enterprise users, including eight of the world’s 10 largest financial institutions. CEO Dr Prince Kohli stated that model choice eliminates vendor lock-in at the AI layer while preserving a single, consistent release assurance system.
