I, like many engineers have a bunch of engineering blogs I like to read every once in a while. Some of these include, Netflix's Engineering blog, Vercel's, Github's and some individual contributor blogs such as Dan Abramov's Overreacted, or Josh Comeau's CSS blog (his blog is a 10/10 if you're like me and dislike CSS).

As much as I like reading these blogs, one thing that's annoyed me is the fact that I have to go to each blog website to see if it's updated. I wanted to be able to "follow" my favorite blogs, much like one follows people on social media. Sure, I can use an RSS Reader, but what's the fun in that!? Naturally, I decided to build my own thing.

This annoyance of mine, led me to build what I call Argus (affectionately named after the hundred-eyed mythological creature). It's basically an RSS Reader for my favorite blogs with a bit of added functionality in the form of AI summaries for each blog posts.

Here's how it works:

• The "sources" list is just some entries within Supabase. Luckily a bunch of blogs have an RSS feed (try adding /feed, or /atom at the end of your favorite blog) so I just saved those URL's within Supabase.
• There's a Cron job that runs on Vercel's platform once a day (to stay within the free tier). This cron job iterates through each of the sources and inserts only new entries into the database. The only thing I'm storing for each "post" is, the URL, the date it was published (if available), the AI summary for the post, and a few descriptive tags.
• OpenAI's API is being used to create the summaries for each of the posts as well as the tags.

It's a fairly simple set up but it allowed me to get up and running within a few hours. I have to say, that although I'm not a big fan around the AI discourse, I do enjoy how fast it has allowed me to move when creating these little side projects.

You can see the project here.