It all depends on what you are doing with RSS. If you subscribe to feeds who barf out 1000 new articles every hour, then a proper filter-system is a basic necce

It all depends on what you are doing with RSS. If you subscribe to feeds who barf out 1000 new articles every hour, then a proper filter-system is a basic neccessary, and machine learning can help there. But I don't know whether this reader is using it for this. Reading the github, it's seems also kinda questionable why they use multiple databases and services. That gives the impressioen that selfhosting this reader will be a pain. I guess I stay with Tiny Tiny RSS (https://tt-rss.org/ https://tt-rss.org/) for the moment.