Inspiration

I've always been a huge read-it-later fan. And I've never been good at keeping a notes system in place. So I decided to build an app in the intersection of those two realms - a read-it-later app with a focus on archiving and keeping articles, and a knowledge base app that stores notes attached to your articles.

What it does

Sink is a read-it-later app first, so you save any articles that seem interesting into Sink, then when you have time you open it up and check the unreads in your Inbox. While reading you can take notes right in the app, and when you are done you archive it. Any time you want to bring it back up, use the full text search to find articles or notes you have saved. And if you don't have time to read a long article - just use the AI Summarizer to quickly get the gist of the article.

How I built it

I built Sink in SwiftUI for the front end, and Supabase as the backend. It was my first time using Supabase for a backend, and so far it seems great. RevenueCat is used to handle subscription statuses, and it is used alongside Superwall for a fully customizable paywall stack. ChatGPT is used for the AI Summary feature.

Challenges I ran into

Deciding defaults - there were so many small decisions to make on an app that is so customizable. Many users will never customize the app, so the defaults I choose are essentially going to be the primary experience of the app for most users. That and deciding what needs to go into the onboarding process. Should it be long, short, should customization happen there, should it be skippable, etc.

Accomplishments that I'm proud of

Sink uses SQLite for a locally searchable database of articles. All data is stored offline, so you can read your articles on a plane or anytime you find yourself without an internet connection.

What I learned

Always submit a few days early, because you never know what detail App Review will pick up on to reject you for.

What's next for Sink - Save Everything

The next big thing is more platforms - a browser extension is coming soon to make saving super easy from any device. That plus native support for both Mac & Apple Vision Pro. I also want to implement more AI features like auto-categorization.

Built With

Share this project:

Updates