Yes. Most organisations build on commercial models rather than training their own, so we focus on how you’ve implemented and integrated that model, your prompts, data connections, guardrails, and permissions, which is where the risks you can actually control live.