Courses

A course in FeynmanLM is a structured collection of related sources — book chapters, PDFs, and video lecture notes — all tagged and decked together. This lets you build a comprehensive knowledge base from a course rather than treating each source in isolation.

Course Structure

Courses are assembled from three types of content:

Content type	How to add
Book chapters	PDF per chapter in FeynmanLM/Courses/<Course Name>/
Lecture PDFs / slides	PDF per lecture in FeynmanLM/Courses/<Course Name>/
Video transcripts	Text file per video in a Podcasts or course folder

There's no formal "Course" object in the app — instead, you use a consistent tag and deck name to tie all course content together.

Setting Up a Course

1. Create a course folder in iCloud

~/Library/Mobile Documents/com~apple~CloudDocs/Documents/FeynmanLM/Courses/<Course Name>/

For example:

Courses/
└── CS229 Machine Learning/
    ├── lecture-01-intro-supervised.pdf
    ├── lecture-02-linear-regression.pdf
    ├── lecture-03-locally-weighted-regression.pdf
    └── problem-set-1.pdf

2. Add lecture PDFs

Copy lecture slides, problem sets, and reading PDFs into the course folder. The app scans this folder and shows each PDF as a source in the Studio sidebar.

3. Generate flashcards lecture by lecture

Select each lecture PDF in the Studio, click Generate Flashcards, and review the proposals. This is best done as you watch or after each lecture — not all at once at the end of the course.

4. Tag cards with the course name

When reviewing proposals, add a consistent tag like cs229 or machine-learning-course to all cards from this course. This makes it easy to filter your review deck by course and track progress per topic.

In the app, you can bulk-tag proposals from a source using the Source batch actions in the Proposal Queue tab.

5. Use a sub-deck (optional)

If you want to keep course cards separate from other books and articles, you can create a sub-deck:

• Navigate to the deck picker in the Create tab or Proposal Queue
• Cards from course materials can be routed to Books > CS229 for example

This is optional — tags alone are sufficient for most use cases.

Adding Video Content

For lecture videos, the workflow depends on the source:

YouTube lectures with captions:

1. Copy the video transcript (auto-captions or provided transcript) into a .txt file
2. Save it to ~/Library/Mobile Documents/com~apple~CloudDocs/Documents/FeynmanLM/Podcasts/<Course Name>/
3. The app treats it as a podcast transcript and shows it in the Studio queue

Coursera / edX videos: Download the transcript from the platform (most courses provide .txt or .srt subtitles) and save it to the Podcasts folder following the same pattern.

Apple Podcasts lecture series: If the course is published as a podcast (e.g., many Stanford and MIT courses are available as audio), add it in Apple Podcasts, listen to each episode, and click Show Transcript in the Podcasts app at least once per episode. FeynmanLM then picks up the cached transcript automatically.

Tracking Course Progress

The Studio sidebar shows how many flashcards already exist for each source. Use this to:

• Identify lectures you've processed vs. ones still pending
• Avoid duplicating cards for sources you've already covered
• Focus your review sessions on the lectures you found hardest

Tips

• Lecture notes > slides: If you have both lecture slides and detailed notes, the notes usually generate better cards (more prose, more context).
• Don't process the entire course upfront. Generate flashcards lecture by lecture, as you watch, so the material is fresh when you review proposals.
• Review within 24 hours of watching a lecture for maximum retention benefit.
• Problem sets: Add problem set PDFs as sources too. The AI generates cards that ask you to reason through the problem-solving approach, not just recall facts.