What tone of voice means for AI-generated text
Tone of voice describes the characteristic way a message sounds: how formal or casual it is, how warm or direct, how long the sentences run, and which words recur. Humans do this instinctively — a lawyer and a summer-camp organiser write the same request very differently. For an AI writing tool, tone of voice is the set of stylistic patterns it tries to reproduce so that generated text matches an intended author rather than sounding generic.
Why it matters
Generic AI text is easy to spot: it is polite, evenly structured, and impersonal. In email that is a problem, because a reply that does not sound like the sender feels off to the recipient and often has to be rewritten. Matching tone of voice keeps a person's or a team's correspondence recognisable and trustworthy, and it removes most of the editing that otherwise cancels out the time an assistant is meant to save.
How it works in DraftKite
DraftKite learns tone of voice from a mailbox's own previously sent messages rather than from a preset persona. When an inbox is connected, it analyses how that person already writes — greeting style, sentence length, formality, sign-offs — and drafts replies in that voice. Each connected inbox keeps its own learned tone, and every draft waits in the drafts folder for the user to read, adjust, and send; DraftKite never sends automatically.