A UK-based startup called Relax.ai is offering localized LLM inference as a service, aiming to give organizations control over where and how their AI models run. The move addresses growing demand from government and enterprise clients who need data residency guarantees and want to avoid routing sensitive queries through US-based cloud providers.

The concept of sovereign inference has gained traction as regulations tighten around AI transparency and data sovereignty. Relax.ai's documentation outlines an approach to running large language models on infrastructure that stays within UK jurisdiction, reducing dependency on offshore providers.

With the Hacker News community already weighing in on the technical and policy implications, Relax.ai represents a practical answer to a question many regulated industries are asking: how do you deploy capable AI models without surrendering control over your data pipeline.