How to Turn One Video Into a Week of Content With AI

You recorded the video. You went live. You had the good conversation on the podcast. The idea already exists, and it's already good. So why does turning it into a week of posts feel like starting from scratch every single time?

Because the hard part was never the idea. It's the reshaping, taking one thirty-minute thing and cutting it into a carousel, three social posts, an email, and a set of talking points, each in a slightly different shape for a slightly different platform. That's repetitive, format-heavy work. Which is to say: it's exactly what AI is good at.

The shift: stop creating, start repurposing

Most people use AI to make content from nothing. That's the weakest way to use it, because a blank prompt gives you generic output. The strongest way is to feed it something you've already made and let it multiply that.

You're not asking the AI to have ideas. You're asking it to do the reformatting you'd otherwise spend a whole afternoon on. That's the whole trick.

The workflow, step by step

Here's the process I run on my own videos.

1. Hand the AI the video. Give it the YouTube link or the transcript. Ask it to pull out the strongest three or four ideas, the ones that stand on their own.
2. Turn each idea into its native format. A key-takeaways carousel. A how-to framework as an infographic. A punchy text post. Each one shaped for where it'll live.
3. Write the captions with it. Hooks and calls-to-action for each asset, in your voice if you've set that up.
4. Get a posting schedule. Ask it what to post Monday, Wednesday and Friday, so you're never staring at a blank week.

One input. A week of output. The part that used to eat your afternoon now takes minutes, and you spend your time on the bit only you can do: deciding what's worth saying in the first place.

Here's what that looks like in practice: three slides from a ten-slide carousel this exact workflow produced from a single video.

The one thing that makes or breaks it

Voice. If the AI writes in generic-AI-speak, readers feel it and the whole thing falls flat. The fix is to give it your real voice to work from: a few examples of posts you've written and liked, and a short profile of how you sound. Do that once and everything it produces sounds like you, not like a robot.

If you want the memory side of this handled properly, here's how to give your AI a permanent memory so you never re-explain your voice again.

Skip the setup: I built it as a skill

Rather than piece this together yourself, I turned the whole workflow into a free bundle you can install into Claude. You give it one YouTube link, and it returns the carousels, infographics and captions, ready to post. It runs with no paid tools to get started.

Grab the Video Repurpose Kit here. It's free, and it's the exact skill I use on my own content.

FAQ

How do I turn one video into multiple pieces of content?

Give AI the video (a YouTube link or the transcript), ask it to pull out the three or four strongest ideas, then have it reshape each into its native format: carousels, infographics, short posts, an email. You're not asking it to invent ideas, just to reformat the ones already in your video.

Will AI-repurposed content still sound like me?

Only if you feed it your voice. Give it a few examples of posts you've written and liked, plus a short note on how you sound, and everything it produces will match. Skip that step and it defaults to generic AI-speak, which readers spot instantly. Setting up AI memory handles this once so you never re-explain it.

What content can I make from a single video?

Realistically: two carousels (key takeaways and a how-to framework), a few infographics, platform-native captions for LinkedIn and Instagram, and an email, plus a posting schedule. That's a full week from one recording.

Do I need paid tools to repurpose content with AI?

No. The Video Repurpose Kit I built runs with no paid APIs to get started. You can add a free key for nicer AI infographics, but without it the kit falls back to clean templates, so it works either way.