Write LinkedIn posts that are actually worth reading.

Yes, if you give it enough of your writing to learn from. Cadiance studies how you reason, how you build an argument, and the specific words and phrases you actually use — then improves with every edit you make.

Fair question, and worth answering directly. You could upload some of your writing to ChatGPT or Claude and ask them to draft a LinkedIn post in your voice. For one post, in one session, you'd get something reasonable. The problems start on post two. General LLMs don't remember you between sessions — every time, you're re-uploading, re-explaining, re-prompting. They don't learn from your edits. They don't know what makes a LinkedIn post actually work today — they're pattern-matching from a training set that includes millions of bad ones. Cadiance builds a persistent voice profile that gets refined every time you edit a draft, grounds posts in your actual documents, and applies specific engineering around what LinkedIn's algorithm and audience currently reward. Yes, we use large language models under the hood — including Claude. But the value isn't the model, it's everything built around it.

Yes. Every edit you make to a draft is remembered, and the next draft lands closer to how you'd have written it yourself. The tool gets more like you, not less, over time.

The specific ones — content where your personality actually shows. LinkedIn posts you've already published are the strongest starting point, followed by longer-form writing like emails to clients, memos, or notes from a talk. The more your samples sound like you thinking out loud, the better the voice profile. Generic content — press releases, marketing copy, drafts someone else wrote for you — is less useful because it isn't actually your voice.

About ten minutes. You upload some writing samples, Cadiance analyses them, and your voice profile is ready. After that, generating a draft takes about ninety seconds — most of it is you deciding whether to publish it, not Cadiance building it.

Yes. Cadiance publishes through LinkedIn's official publishing API — the same one LinkedIn provides to its own partners for scheduled posting. We don't automate likes, we don't send connection requests on your behalf, we don't scrape your data, and we don't run scripts against LinkedIn's interface. LinkedIn accounts get restricted for behaviour that pretends to be human when it isn't. Cadiance drafts posts — you decide when and whether to publish them.

Two separate things, both important.

Your LinkedIn account: Cadiance connects through LinkedIn's official publishing API, using OAuth. You authorise the connection, you can revoke it at any time from your LinkedIn settings, and Cadiance never sees or stores your LinkedIn password.

Your writing samples: they're used only to train your voice profile, they're stored securely, and they aren't used to train models for any other user. Your voice profile is yours, permanently bound to your account. If you cancel, you can request deletion of everything.

Yes. In Cadiance for Teams, a comms manager or admin can draft, schedule, and publish posts on behalf of each executive on the team — and the posts appear under the executive's own LinkedIn account, not the comms manager's. Each executive connects their LinkedIn to Cadiance once. From there, the comms team can work in each of their voices, run drafts through approval, and schedule content into each person's feed. Nothing is shared. Nothing crosses over. Every voice, every draft, every approval trail stays inside your workspace.

LinkedIn only. We built Cadiance specifically for the way professional voice works on LinkedIn — the format, the audience, the algorithmic dynamics. Trying to be a universal cross-platform tool would mean being average at all of them. We'd rather be exceptional at one.