Khaiso
For quite some time, I have been obsessive with turning semi-structured data to structured data. So, in 2022, I decided to create a recipe parser. While my initial parser was horrendous, after iteratively improving it on and off over the course of a month, the scraper became surprisingly good. It's capable of extracting recipes from sites lacking metadata, like JSON-LD, and preserves ingredient and direction organizational groupings if present. This made the scraper leagues ahead of the competition, like Samsung Food (formerly Whisk). So, I thought it might be cool to build a recipe app to put the scraper to use. I've been working on the mobile and web app on and off for quite a while.
I've made quite a lot of changes to the tech stack, most of which resulted from me wanting to try an new framework or library. I started off with a Firebase backend, but the cold starts times felt unbearable. Even though, at scale, cold starts would be more negligible, I decided to swap to AWS, mainly because I wanted to learn it. Khaiso helped me pick up architecting AWS services with infrastructure of code via AWS CDK. To date, I've learned around 10 services just through Khaiso.
Even though using Next.js was buttery smooth, I also swapped it out for SvelteKit. I really wanted to learn SvelteKit because it seemed very promising. I ended up having to write my own UI library, but I got to learn accessibility and web ARIA best practices as a result.
Let me know if you'd like to beta test the app. I'm always looking for more testers.
