As we know, AI has widened the gap between how fast code is written and how fast it can be verified. Seemingly outlandish (but probably accurate) figures estimate that developers can now produce 741% more code, but actual release velocity has risen by under 20%. Test automation company Sauce Labs wants to close the gaps that exist in the automated coding space with bring-your-own-model capabilities for AURA, its AI-Unified Release Assurance platform.
AURA’s new bring-your-own-model capabilities are designed to allow software teams to develop tools and services using their choice of open source, open-weight, or proprietary LLMs. Users will also have the flexibility to change models as technology, regulations, or enterprise standards evolve, all while preserving one consistent system for verifying software.
The CEO known as Prince
CEO of Sauce Labs. Dr. Prince Kohli has said that these moves standardise release assurance, without forcing organisations to use leverage a single AI model provider for coding.
“Enterprise AI should expand choice, not create a new layer of lock-in,” said Dr. Prince Kohli, CEO of Sauce Labs. “AURA lets users choose the intelligence layer that fits their business while keeping one consistent, governed system for release assurance. Models can change; the foundation for production confidence should not.”
The company suggests that AURA “closes the loop” from business intent to production confidence by authoring, executing, analysing and self-healing tests at AI speed, with humans in control.
Changing models, solidified assurance intelligence
An expanded architecture separates the model layer from Sauce Labs’ release assurance intelligence, agentic workflows, and execution infrastructure. Software developers now have flexibility to change models as technology, regulations, or enterprise standards evolve while preserving one consistent system for verifying software.
Enterprises running AURA achieve independently validated results: 90%+ fewer production incidents and 47% faster release cycles with 38% of engineering capacity reclaimed.
Model choice is now a strategic requirement
As enterprises move from experimenting with AI code to deploying it in governed, production-scale environments, Kohli and team insist that “model choice becomes a strategic requirement” and no single model will be the best fit for every organisation or moment in time.
AURA gives developers the flexibility to choose supported models that align with internal security, privacy, governance, and architecture standards. Engineers can evaluate and change models based on accuracy, latency, cost, and policy without replacing the release assurance platform.
This flexibility allows enterprises to mix and match models for their needs while preserving consistent confidence for release assurance.
Don’t mess with context
“Some vendors now acquire context engines and sell enterprise context back as a platform feature. Sauce Labs takes the opposite position: context belongs to the customer. Enterprises have already built it, in the models they have selected, tuned, and governed, and in the systems and data those models draw on. AURA connects to that investment instead of replacing it,” states the company.
Informed by more than 8.7 billion test executions, AURA verifies how software behaves under real conditions, not just on the basis of how repositories and tickets say it should behave.
The loop does not stop at release. AURA captures production errors and feeds them back to refine business intent and sharpen the next round of tests, so release quality improves with every cycle. Customers keep ownership of their context, choose their models, and run one governed system for release assurance.
“You cannot buy a customer’s context and sell it back to them,” said Kohli “Enterprises already own their context: their models, their systems, their data. AURA respects that investment. Customers bring the intelligence they trust, and we bring the verification layer that proves software is ready to release.”
Open by design
The model is only one component of reliable release assurance. AURA pairs the selected LLM with Sauce Labs’ domain intelligence, agentic workflows, execution across more than 10,000 devices, and a learning loop informed by more than 8.7 billion test executions. Enterprises gain model freedom without losing the context and infrastructure required to determine whether software is ready to release.
The announcement extends Sauce Labs’ long-standing commitment to open ecosystems. Created by the founders of Selenium and Appium, Sauce Labs applies the same principle to enterprise AI: customers should be able to adopt the tools and models that serve them best while relying on a common platform for confidence at scale.
Because AURA is framework-agnostic, CI/CD-native and designed for existing development environments, organisations can modernise the model layer without a rip-and-replace migration or fragmented release controls.
AURA’s model choice capabilities are available now for Sauce Labs enterprise customers.