ThoughtRipple

I was the first user. I kept a few questions nearby so the product stayed about writing.

QWhat does someone need to finish a post today?

AAn account, an editor, a cover if they want one, and a way to publish.

QWhere does AI help?

AA short description can become a draft. Chat is there to think out loud. The voice stays yours.

QWhat should feel easy to trust?

ASigning in, billing, and the first welcome email. That’s why accounts, payments, and mail are part of the product.

I wrote a handful of stories so I could try a few directions before picking one.

• As a writer, I want to create an account and land in a dashboard so I can start a post.
• As a writer, I want a blank editor, or a draft from a short description, so I can get moving.
• As a reader, I want a clean public post with a shareable link.
• As a writer, I want a free plan that publishes, and a paid plan if I want AI and more posts.

Publish, and it becomes a real page — cover, byline, a link. You can put your name on it, or stay anonymous.

The app needed a real backend, not a mock. I created a Supabase account first, then linked it to Cursor. Auth for accounts, tables for profiles and posts, storage for covers, and row-level security so a writer only sees their own work. The schema lives as migrations in the repo, so the database grew with the product instead of living only in a dashboard.

Paid plans go through Stripe. Starter and Pro, a hosted checkout, then back to the dashboard. That part is straightforward. Stripe doesn’t write to the app on its own, though. I set up a webhook: when someone subscribes, upgrades, or cancels, Stripe sends an event to a Supabase Edge Function. The function checks the signature, then writes the subscription into Supabase so the dashboard already knows the plan when they land.

• A new checkout writes the plan into Supabase.
• An upgrade, renewal, or failed payment updates it.
• A cancel lands too, so the plan you see is the plan you have.

People looked for the price before they looked for features. The home page landed better once the promise sat up top and real posts weren’t far behind. Opening a live story was the moment it clicked — that’s when it felt like a place, not a pitch.

Almost everyone opened it on a phone first. Sign-in needed to be easy to find. Long pages were fine, as long as the next step stayed close — a story, a plan, a way into the desk.

Writing on a phone is a different mood. The menu, the editor, chat — they feel better stacked than squeezed. New Post had to be the obvious move, or the desk went quiet.

ThoughtRipple is live at thoughtripple.blog. Come sit down and write.

This is the first product I designed and built all the way through — from a feeling to a live site. I did the work in Cursor: describe the next piece, try it, keep going. That’s where I learned what UX engineering actually is. Not just screens. How the screens, the account, and the thing people open in a browser stay one product.

I was the project manager too. Solo doesn’t mean there’s no one holding the thread. It meant deciding what to build next, when it was good enough to show someone, and when to ship. That part taught me as much as the pixels.

Along the way I found tools I hadn’t lived in before. Supabase for accounts and posts. GitHub for the code. Stripe for plans. I didn’t know the insides when I started. I do now — enough to wire them, enough to keep building.

• How to sit in Cursor and actually ship.
• How to run a project when you’re the only one in the room.
• The ins and outs of UX engineering — auth, a webhook, a layout that works in a pocket.
• How to take a feeling and make a home people can write in.