FAQ

You build the brand's language into the AI itself, rather than prompting a general tool each time. When voice, knowledge, content structures and governance are encoded into the environment where content is created, every output resolves through the same source – so on-brand stops being a review step and becomes structural. That's what Arcroyd is: your brand's own AI, built from your brand rather than prompted toward it.

Because general AI has read everyone's language and holds no one's. Prompted for your brand, it produces something plausible but generically averaged – close, but never quite yours – and the more you create, the further the voice drifts. Drift isn't a writing failure; it's the absence of a governing system. Guidelines describe how a brand should sound but don't govern how it creates, and at AI scale that gap widens fast.

Guidelines were written for people to interpret, and they sit outside the work. AI doesn't open the tone-of-voice document; it generates from whatever structure it's been given, usually a short prompt assembled from memory. So the brand a guideline describes and the brand an AI produces steadily diverge. What holds at scale isn't a better document – it's a system that governs language at the point it's created.

Writer, Jasper and similar platforms are shared products built to automate content at scale, increasingly through agents. Your brand is one input among many: voice profiles, terms and style rules configured inside someone else's platform, alongside every other customer's. Arcroyd is built the other way round. Your language is first defined and documented as a full system you own, held outside any tool. It is then encoded into one environment, built and operated for your brand alone. A system you own, not a setting you configure.

Writer, Jasper and similar platforms are shared products: one application serving thousands of brands, each configuring a voice profile and a word list inside it. Your brand lives as a setting in someone else's tool, alongside everyone else's. Arcroyd is the opposite by design – one environment, custom-built and operated for a single brand, governed by your full documented language system, encoded in depth and owned outside the tool. A system you own, not a setting you configure.

They average toward whatever you feed them – inconsistencies and drift included – and the voice they learn is a statistical echo, not a governed definition. Arcroyd works the other way: your language is architected and documented as a system first – voice, knowledge, structures, rules, governance – then encoded. It creates from the brand as it's defined to be, not a read of what it has been. And that definition is yours, held outside the tool.

Most tools in this space are self-serve SaaS. You buy a licence, and your team configures the brand, keeps it current and carries the work when the person who built it moves on, with vendor support or without it. Arcroyd isn't. It's built, operated and stewarded by Rule of Three: your environment is stood up, encoded, monitored and kept current for you, so brand governance never becomes another system your team has to run. Owned by you, operated by us.

No. Arcroyd sounds like your brand, because it creates from your encoded language system rather than a generic model nudged in your direction. It doesn't sound like AI – it sounds like your brand, down to the detail that usually gives generated content away.

No. A short brief in, on-brand content out – for anyone on your team, writer or not. The expertise sits in the system, so more of your people can produce work that genuinely sounds like your brand, without becoming brand experts themselves.

It stays yours. Your documented brand language system – the asset the engagement produces – is owned by you and held outside the tool, never locked inside it. The runtime that operates Arcroyd is licensed and operated by us; the system beneath it is yours regardless. If you move on, your language system moves with you.