This article is a work in progress log documenting my progress with using Anki as a learning tool in my Software Engineering career.

March 2026

My learning method is very simple:

While reading a book/article I make an Anki card when I find something interesting. Same for when I watch a Conference/Talk.

I ankify cards immediately, rather than hightlight and ankify during the review simply because I don’t trust myself that I will review the book at all. There are so many interesting books/articles to read that I might start to read a new book/article immediately after the current one, rather than review it and ankify the highlighted parts. I still highlight, but only because I might need to reference the highlighted part that I have ankified if the card is not clear in the future and I need to reread that specific section.

I use a hierarchical deck, but when reviewing I review all decks at once to promote interleaving.

I use custom notes with a “Source” field so that it’s easy to go back to the source material if needed. A Source can be a link, book + page combo or YouTube + timestamp combo.

I use Obsidian as a task manager and scratchpad for random notes. I don’t take notes on it for learning purposes anymore because I never review them anyway and I sometimes even forget that I have notes on certain arguments.

My notes are small, connected and meaningfull. I try to make them bidirectional. Sometimes I ask the same question in different modes.

I use images of charts/diagrams as answers when possible. It’s easier to recall them than a plain list, even though I’m not a visual learner.

I add/edit on my computer and I review on my phone whenever I can, be it while waiting or between sets at the gym. I try to review at different places to avoid anchoring my memory to a certain location. I sometimes review on my computer for the same reason.

I’m always editing cards, to split them, join them, clarify them. But I don’t stress too much about it when creating a card. It’s better to add a card and improve it later than to try and make it perfect from the start. You need the insight you gain from trying to answer to improve the question itself. Sometimes a question is too revealing (it has too many hints for the anwer) and sometimes it’s too obscure (the answer could be one of many).