How do you write in an executive's own voice with AI?
Most companies want their founder or CEO to publish regularly under their own name. People trust a person more than a logo, and a founder explaining why they made a hard call gets read in a way a company announcement rarely does. The usual answer is a ghostwriter, whether that's someone on the marketing team, an agency, or a chatbot.
The trouble is that ghostwritten drafts often don't sound like the person. The executive reads the draft, says "I'd never say that," and rewrites half of it. After a few rounds, they stop asking for drafts and stop posting. The ghostwriter usually had good intentions and a description of the voice, such as "confident, warm, and direct." But a description like that fits thousands of executives, so it can't tell a writer how this particular person sounds.
The fix is to learn the voice from the person's own writing, then write the habits down with evidence. One CEO states the point in a single blunt line, then tells the story of the customer who taught it to her. Another writes long, reflective paragraphs and ends with an open question. A third never uses emoji and never opens with a question. Those are habits a writer can follow and an editor can check.
An AI agent is well suited to this work. It can read a year of someone's posts in minutes, quote the lines that show each habit, and apply the same profile to every draft without drifting back toward its own default style. It can also keep track of every story the executive has told, so it doesn't ask for the same one twice.
Kite is an AI marketer. It researches your business, builds a growth strategy, and ships the work that brings you new customers. Here's how it learns an executive's voice and writes in it.
How Kite writes in an executive's own voice
Kite starts by collecting the executive's real writing. With a LinkedIn profile, it reads the public profile and the person's recent original posts. Without one, it works from a short interview and any samples you share, such as an old essay, a newsletter, or a transcript of a talk.
From that writing, Kite builds a voice profile on the person's own page in your team's knowledge base. The page records how they write, what their writing never does, the themes they want to be known for, and the stories and numbers they've shared. Whenever Kite writes something under that person's name, it writes from that page.
The executive stays the author. Kite drafts, the executive approves, and for LinkedIn posts, the executive posts from their own profile. Every edit they make on a LinkedIn draft gives Kite more evidence about how they write.
- How do you create a writing voice profile from someone's past posts?
- Record what their writing never does
- Collect the stories and opinions only the executive has
- Write everything under their name from the same profile
- Approve the draft, then post it yourself
- Learn from every edit the executive makes
How do you create a writing voice profile from someone's past posts?
A useful voice profile describes habits and proves each one with a quote. When an executive shares their LinkedIn profile, Kite reads it along with up to ten of their recent original posts, skipping reposts, because a repost shows what someone likes rather than how they write. Kite then looks for the traits that show up in that writing and records each one next to a line the person wrote, with where and when they wrote it.
A trait Kite can't back with a quote stays off the page. This rule keeps the profile honest. "Uses short paragraphs" is a guess until there's a post that shows it, and a guess in the profile becomes a mistake in every draft.
Here's what one entry might look like:
- Leads with a conviction, then shows the reasoning. "Most onboarding problems aren't design problems. They start in the sales call." (LinkedIn post, March 2026)
Kite also checks that the profile belongs to the right person. If the LinkedIn profile it reads carries a different name from the person setting up, Kite asks before using anything from it, so one person's voice never ends up on someone else's page.
Record what their writing never does
What a person never does is often the clearest part of their voice. Kite records these absences in a separate section of the profile, and it only counts a pattern that holds across at least three of the person's posts. One executive never opens with a question. Another never uses emoji, and a third never ends a post by asking for comments.
The executive can add to this list directly. If they read a draft and tell Kite, "That's not how I write," Kite moves that trait from the list of things they do to the list of things they don't, in their own words. When the rule applies to every piece of writing, such as "never use emoji," Kite also saves it as a standing preference for the team.
This section is what keeps drafts from sliding back into the generic style most AI writing has. A writer who knows the executive never opens with "Here's the thing" won't open with it.
Collect the stories and opinions only the executive has
A voice is only half of sounding like someone. The other half is the material: the customer conversation from last Tuesday, the number that surprised them, the industry habit they think is wrong. A post built on that material sounds like the person, because it comes from the person.
For LinkedIn posts, Kite checks for three things in the executive's own words before it drafts:
- A specific moment. "On Thursday's call, our biggest customer asked us to remove a feature" is a moment. "Customers value simplicity" is a theme.
- One concrete detail. This is a number with what it measures and over what period, or an exact before and after.
- Their point. That could be what changed their mind, who would disagree with them, or what the reader should do differently.
When something is missing, Kite asks one to three short questions in a single message. If the executive says "just draft it," Kite writes without the missing piece rather than inventing one, because a guessed number or opinion would go out under their name.
Every answer is saved word for word in the person's story bank on their page. Kite keeps the executive's strongest phrases exactly as they said them, and each new draft can draw on everything they've told it before.
Write everything under their name from the same profile
Once the profile exists, Kite uses it for any writing that carries that person's name. That includes LinkedIn posts, thought-leadership articles, a founder's note in a newsletter, an email that goes out from the CEO, or a talk abstract for a conference. Your company's brand voice stays on its own page, so a company blog post and a founder's essay don't blur together.
Each executive gets their own page. Your CEO and your head of product can both have profiles, and Kite writes each piece from the right one. A marketer on the team can ask for a draft in any executive's voice. For example, you might write, "Draft a post for Priya about the customer story from Thursday's all-hands." Kite builds it from Priya's profile and story bank.
For LinkedIn, Kite can also run a weekly rhythm for each executive. Every Monday morning, it sends them the week's drafts, a few timely topics they could take a position on, and one question about something that happened at the company. Each draft shows which of the executive's own words it was built from.
Approve the draft, then post it yourself
A post on an executive's personal profile should come from that executive, so only the author can approve it. Teammates can see the drafts and ask for changes, but the approval belongs to the person whose name is on the post. If anyone changes a draft after it was approved, the approval clears and the post goes back to the author.
Once a LinkedIn post is approved, the executive copies it to LinkedIn, posts it from their own profile, and pastes the link back into Kite. For other writing, Kite asks for approval before it publishes a page or sends an email, just as it does for any of its work.
Learn from every edit the executive makes
For LinkedIn posts, each round of edits teaches Kite something about the executive's voice. Every week, Kite compares each draft it wrote with the version the executive approved, and it reads the reasons they gave for any drafts they skipped. When a pattern shows up across two or more posts, such as "shorter openings and no closing questions," Kite records it on the executive's profile and applies it to that week's drafts. One edit on its own isn't a pattern, so Kite waits for the second before changing the profile.
The next weekly plan opens with a short note on what Kite changed. Over time, the drafts need less editing, and the executive spends their time on the stories and opinions rather than on fixing sentences.
Kite also reads how each LinkedIn post performs after a day and after a week. It tells the executive how the post compares with their usual results, so they can see which kinds of posts their audience responds to.
What you get
A voice profile for each executive
A page in your knowledge base that describes how they write, with a quoted line of theirs behind every trait.
A list of what they never do
The openings, phrases, and habits their writing avoids, drawn from their posts and their own corrections.
A story bank
Their stories, numbers, and stated positions, saved word for word so nobody has to ask for them twice.
Drafts in their voice
LinkedIn posts, articles, newsletter notes, and emails written from their profile and their own material.
A profile that sharpens with use
Patterns from their edits on LinkedIn drafts are added to the profile, and each weekly plan says what changed.
How to get started
Start with one executive. You can work with Kite in your browser or in Slack. The executive can open the Social tab in the Kite web app and choose to set up their LinkedIn posts, or simply tell Kite, "Set up my LinkedIn." Kite asks one question first: what happened at work lately that they'd tell a smart friend. While they answer, it reads their public LinkedIn profile and recent posts.
After a follow-up question or two, Kite writes their voice profile, proposes two or three themes and a posting rhythm, and saves the first drafts for them to review. The executive needs to do this setup themselves, because the voice and the stories have to come from them.
If the executive doesn't post on LinkedIn, share other writing instead. Paste a few samples into the chat, or upload an essay or a talk transcript in the web app, and ask Kite to write the next piece in their voice. The more of their real writing Kite can read, the closer its first draft will be.
You don't need to connect anything. Kite reads public LinkedIn profiles and posts with its own built-in research.
Integrations used to write in an executive's own voice
Kite does this work with its own built-in tools. It reads the executive's public writing with built-in research, keeps the voice profile in your team's knowledge base, and saves LinkedIn drafts in Kite for the executive to approve.
Learn the voice from their posts
Kite reads the executive's public profile and up to ten of their recent original posts to build the voice profile, then reads the reactions, comments, and reposts on each published post after a day and a week.
Common questions
How many writing samples does AI need to learn someone's voice?
A handful of real pieces is enough to start, and more is better. Kite reads up to ten recent LinkedIn posts, and it needs a habit to show up in at least three of them before it records that habit as something the person never does. A couple of sentences can show a person's vocabulary, but they can't show how that person builds an argument or ends a post. If the executive has never posted, Kite starts from their interview answers and the samples you share.
Is this the same as AI voice cloning?
No. Voice cloning copies the sound of someone's speaking voice for audio. A writing voice profile describes how someone writes: their sentence length, their openings, their favorite moves, and what they avoid. Kite's voice profile is about writing, and it uses the profile to draft text that reads like the person.
How do I train AI to write like me?
Give it your own writing rather than a description of your style. Share your recent LinkedIn posts, a few essays or emails you're proud of, or a transcript of you talking through an idea. Then correct the drafts honestly. With Kite, you can say "that's not how I write" about any trait, and it moves that trait to your list of things to avoid. On LinkedIn drafts, Kite also compares your edits with what it wrote and adds repeated patterns to your profile.
Can one person manage writing for several executives?
Yes. Each executive has their own voice profile, story bank, and themes, so a CEO's posts and a head of product's posts stay distinct. A marketer or chief of staff can ask Kite for drafts in any executive's voice and follow their weekly plans. Each executive still approves their own LinkedIn posts.
Can Kite post to the executive's LinkedIn for them?
Kite drafts the posts and the executive posts them. Once a post is approved, Kite gives back the final text for the executive to paste into LinkedIn, and the executive pastes the post's link back into Kite so it can follow the results.
Is it okay to ghostwrite for a CEO?
Yes, as long as the CEO holds every opinion in the piece and approves it before it goes out. Executives have long worked with ghostwriters and communications teams, and the standard has always been that the person named stands behind the words. Kite works to that standard by building each draft from the executive's own stories and positions and showing which of their words each LinkedIn draft used.