I took this project down, temporarily, for improvements. Previously, this was actually a back-end product for a university so that students could search multiple journals simultaneously (something they still cannot do). It required a lot of API logins, and it was very difficult to administrate.

That being said, it was an incredibly useful product! Students saved hours by being able to use EXACT PHRASE and {journal A} -OR- {journal N}. The goal is not only speed: the idea is that we should be able to search the top 20 ABA journals and know that we have exhaustively discovered all articles matching exact search terms.

The reason this product is so difficult to develop is because it relies on some joins that were taken out of Google, and a lot of universities just rely on Google; so, I think of this as a search engine.

This is my explanation of why it is labeled as an AI product: The new features will include the following: (1) Save files, notes, completed assignments, and canonical case examples study-by-study (similar to Zotero). (2) Additional -OR- search capability for things like author and title (also not available on Google anymore, sadly). (3) Command-line arguments allowing the app's main functions to be used in non-visual/text-only environments. (4) Web-pipes to create HTML dashboards for each study (a pipe is a tool that updates the HTML when new information is gathered via a CLI). (5) The system can be connected to SQLite, PostGRE, MariaDB, or QXML so that high-compliance medical offices can succeed during audits (must use always-offline installer) so that identifying info is unable to contact the internet.

My goal in rewriting before release is primarily just to create a FOSS product with good isolation so small medical offices will be able to contribute to the knowledge base without paying punitive taxes.