Thanks for raising this, Dylan Leach — totally hear you on how a blunt “no” (or negatively coded phrasing like the examples you shared) can create a rough experience. To help us pinpoint where to improve: where in DriveCentric are you seeing this come up most (customer chat, email/text generation, call summaries, internal notes, etc.)? And when the AI would otherwise say “no,” do you want it to always hand off to an agent, or only for specific topics/situations? If you can also share 2–3 real snippets of responses you’d like to avoid (and what you’d prefer it to say/do instead), that would be super helpful for us to validate the right behavior.