Guidesnap

Post:

Six months ago my wife showed me her workflow for turning screen recordings into documentation. Watch the video, pause, screenshot, type, repeat. Every week.

I went looking for a tool that solved it. Everything either made you re-record inside their extension, or cost like enterprise software. Nothing just took a finished recording and gave back a clean guide.

So I built Guidesnap.

You upload any screen recording, AI transcribes the narration, detects the steps, grabs screenshots at the right moments, and ships a PDF guide in minutes. Edit any step in the browser, no re-uploading.

The build:

Two months, solo. Whisper for transcription, Claude for structuring steps, Modal for video processing, React 19 / TanStack Start on Cloudflare Workers, Supabase for backend, Paddle for billing.

The hardest part wasn't the AI pipeline. It was distribution. I launched on Product Hunt, got 13 real comments, solid feedback, but almost no signups. Classic first-timer mistake: I focused on building, not on building an audience.

What I learned:

PH rewards preparation, not execution. The launches that win have 6 weeks of coming soon pages and warmed audiences. I had two weeks.

The real work starts after launch. Reddit, LinkedIn outreach, SEO, direct conversations with people who have the pain. Slower but it's what actually brings users.

Shipping solo is doable. Knowing who your user is before you ship is not optional.

Where I am now:

30 registered users, product is live, first paying customers coming in. Building in public and figuring out distribution one channel at a time.

Happy to answer anything about the build, the stack, the launch, or what I'd do differently. And if you record walkthroughs and hate writing docs after, try it free: guidesnap.app