From a Book to a Study Plan: How pimReader Actually Teaches

Here is the uncomfortable truth about reading to learn: you forget almost all of it.

You finish a book on Roman history, or a grammar guide, or a textbook chapter, and two months later you can recall the vibe and maybe three facts. The reading felt productive. The retention wasn't there.

pimReader exists to close that gap. Not with more highlighting — highlighting is where good intentions go to die — but with a loop that takes something you read and carries it all the way through to long-term memory.

Let's walk the whole loop once.

Step 1: Collect

Everything starts in your library. EPUB, PDF, TXT, HTML, Markdown. News articles pulled from RSS, Atom, OPDS catalogs, Telegram channels, or X. Web pages saved for offline reading, stripped of banners and popups.

The point isn't hoarding. The point is that everything you want to learn from lives in one searchable place — so the assistant that helps you learn from it can actually see it.

That's the part most AI tools get wrong. A generic chatbot knows the internet. It doesn't know your textbook, your notes, or what you failed to remember last Tuesday.

Step 2: Ask the Mentor

Open Ask AI and describe what you're trying to do. Not "explain photosynthesis" — that's what search engines are for. Something like:

"I want to get through this Spanish grammar book in eight weeks, about 30 minutes a day. I'm roughly A2."

The Mentor asks clarifying questions — your level, your deadline, what you can realistically do per day — and then proposes a study: a sequence of lessons sized to that goal.

Crucially, it proposes. Nothing is created until you approve it. Generated plans show up in a panel where you read them first, and if the plan is too ambitious or too shallow, you say so:

"Too many lessons. Cut it to six and drop the subjunctive for now."

Iterate until it fits. Then approve.

Step 3: The lessons become real material

This is the part that surprises people. A lesson isn't a stub in a list — the Mentor can write the actual content.

From the lesson menu you get:

• Create lesson book — a full book, written into your library, that you read like any other book in pimReader
• Create exercise book — a companion book of practice work
• Run quiz — an interactive quiz over the lesson content
• Open Flashcards — vocabulary or facts from the lesson, ready for repetition
• Vocabulary — the word list the lesson is built around

And because a lesson book is just a book, everything else in pimReader works on it: the dictionary, text-to-speech, bookmarks, search.

If a lesson book isn't quite right, you can ask for a patch — a targeted revision you review and apply, rather than regenerating the whole thing.

Step 4: Turn insight into flashcards

While you read — a lesson book, a novel, a news article, anything — you'll hit things worth keeping. A word. A date. A definition that finally clicked.

Select it. Ask for a flashcard. You get a question and an answer, filed into a folder.

You can also hand the assistant a passage and ask it to propose a set of flashcards, then pick the ones worth keeping from a preview. Tables become flashcards. Vocabulary lists become word cards with the right language attached.

The friction is the whole game here. If making a flashcard takes ninety seconds, you won't. If it takes two taps, you will.

Step 5: FSRS does the remembering

Now the schedule takes over.

pimReader uses FSRS, a modern spaced-repetition algorithm that models how fast you forget each specific item. Not fixed intervals applied uniformly — an actual per-item forgetting curve.

When an item comes up, you rate it: Again, Hard, Good, Easy. Each button shows what it will do — for a new item, something like 5 Min, 1 D, 2 D, 4 D. You can see the consequence of your honesty before you commit to it.

Rate honestly and the schedule does its job: things you know drift out to months, things you don't come back tomorrow.

Step 6: Practice, don't just review

Recognition is not the same as knowing. So the same material can be practised several ways:

• Memorize Words — short sentences that commit vocabulary to memory
• Exercise — build your own sentences from the words you're learning
• Develop Fluency — understanding and composing longer sentences
• Practice Dialogs — a lifelike conversation, line by line
• Practice with AI Assistant — live chat using your own vocabulary
• Speaking practice — an entirely spoken session, hands-free

That last one deserves emphasis: you talk, it listens, it answers, and the results go straight back into the same FSRS schedule. No separate app, no separate progress to track.

Step 7: Watch it move

The Studies view shows where you actually are: which lesson is In progress, which are Not started, which are Completed. Not a streak counter designed to guilt you — just the state of the work.

And it syncs. Start a lesson on a tablet, finish the review queue on a phone during a commute.

The loop, in one line

collect → plan → generate → capture → schedule → practise → track

Every step feeds the next. The book you read on Tuesday produces the flashcards you review on Friday and the conversation you stumble through next month.

None of this is specific to languages, by the way. Swap "vocabulary" for "case law", "drug interactions", or "Kubernetes concepts" and the loop is identical.

That's the design. Start anywhere in it — most people start by asking the Mentor one honest question about something they're already reading.