A typical week as a student involves four or five lectures, a couple of seminars, three or four chapters of reading, problem sets, a study group meeting, and at least one assignment due. By midterms, the notes from week 1 are scattered across a notebook, a Google Doc, two different folders on your laptop, the slide deck the prof posted on the LMS, and a flurry of texts in a study group chat. When you sit down to review, you don't actually have one place to start.

The note-taking problem at school is mostly an organization problem dressed up as a content problem. The lectures aren't that hard to follow in real time. They're hard to find again three weeks later when you actually need them. A vault that holds lecture notes, readings, problem sets, and study-group output in one searchable shape — with the agent doing the synthesis when exams approach — fixes most of it.

## One vault, one parent page per term, child pages per course

In Docapybara, the term gets a top-level page: *Spring 2026*. Underneath sit child pages for each course you're taking. Each course page has child pages for *Lectures*, *Readings*, *Problem sets*, *Study group*, *Office hours*, and *Exam prep*.

Pages nest indefinitely, OneNote-style. So *Spring 2026 > Linear Algebra > Lectures > Week 4 > Tuesday* is a valid path; so is *Spring 2026 > Econ > Readings > [chapter title]*. The agent searches across the whole tree regardless of how deep it goes.

If you're also juggling a job or an extracurricular alongside coursework, the vault holds those too. The *student* category just becomes one branch of a broader vault. (For the broader case, [Why Your Notes App Should Be the Same App for Work and Life](/guides/personal-life/same-app-work-and-life/) covers the argument.)

## Lecture capture — the audio that does most of the work

The single highest-value move for most students is recording lectures (when allowed by the professor). Audio recording in-app, with speaker labels via diarization, gives you a transcript with the professor and any students who asked questions clearly attributed. Tap record, leave it running, hand-take notes if you want, listen the way you actually listen.

After lecture, the transcript lands on the *Lectures > Week N > [Date]* page. You can clean it up later or leave it raw. The agent can summarize: *"Summarize today's lecture — main argument, examples used, anything the prof emphasized."* You get a clean review document.

For the questions students asked, the agent can pull: *"What did students ask in today's lecture, and what did the prof say in response?"* Useful when the question someone asked is the exact thing you're confused about a week later.

For lectures with slide decks, drop the slides on the page as a PDF. Docapybara converts uploaded PDFs to markdown via docstrange, so the agent can read them — *"in slide 17 of Tuesday's lecture, what was the formula for [thing]?"* gets answered with the actual content.

## Readings — turning chapters into searchable text

For the assigned reading, the workflow depends on whether the readings are PDFs, books, or article links. PDFs drop on the *Readings* page directly. The agent reads them as text. *"In the chapter on monetary policy, what was the section on the Taylor rule?"* gets answered with the actual passage.

For physical books, voice notes after each reading session work well. Tap record, walk for two minutes, talk about what stuck. Transcript lands on the book's page. The agent can pull at exam time. (For deeper reading-tracking specifically, the [Tracking Your Reading and Building a Personal Library](/guides/personal-life/reading-notes-personal-library/) shape extends this — the student version is reading-list discipline at higher density.)

For article links and supplementary readings, drop the URL with a one-paragraph summary in your own words. The agent can summarize across them at exam time: *"For the unit on industrial policy, pull every reading I had and the one-paragraph summaries I wrote."*

## Problem sets — the work plus the reasoning

For courses with problem sets (math, CS, econ, physics), the *Problem sets* page tree captures both the work and the reasoning. One child page per problem set, with the problem statements at top, your work in the middle, and a *what I learned / what tripped me up* note at the bottom.

The work itself can be hand-written and photographed (drop the photos on the page) or typed inline. For the *what I learned* notes specifically, the agent can pull at exam time: *"Across the problem sets so far, what were the techniques I struggled with that the prof said would be on the exam?"*

For study group sessions where you worked through problems collaboratively, voice recordings of the conversation can land on the page (with everyone's consent). The agent can pull: *"In the study group on Thursday, how did we end up solving problem 3?"* Comes back with the conversation summary. (The same shape — capturing collaborative thinking sessions — applies to [How to Capture Action Items So They Actually Get Done](/guides/meetings-people/action-items-actually-get-done/) at the meeting level.)

## Study group — the hive memory

Study groups are usually the most underused note-source in a student's vault. The conversations are valuable; nobody writes them down; everyone walks out with a partial copy in their head.

A *Study group* page per course captures the running thread. After each session, drop a summary — voice works well. Who showed up, what topics you covered, what nobody understood, what you decided to look up before next time. The agent can summarize the running themes: *"What topics has the study group kept coming back to this term, and which ones are still confusing?"*

For shared notes within the study group itself, you'll likely use a shared doc your group uses. Docapybara is single-user — your vault is yours. Paste the shared notes into your own vault as well, so the agent can search them alongside your other material. The shared doc stays the working surface; your vault holds the searchable archive.

## Office hours — the highest-leverage notes most students don't take

Office hours conversations are usually the highest-density learning of the week, and the worst-documented. Most students walk in, get the answer, walk out, and forget the framing within a week.

A short voice note after each office-hours visit fixes most of it. *"Office hours with Prof X today on the dual problem in linear programming. The thing I was missing was the geometric intuition — slack variables correspond to constraints that aren't binding. She said this is on the exam."* Two minutes, transcript lands on a *Office hours* page or on the relevant lecture page.

The agent can summarize at exam time: *"Across office hours visits this term, what did the prof emphasize as exam-relevant?"*

## Exam review — where the vault earns its keep

Exam prep is the moment when the vault stops being a habit and starts being a tool. Two or three weeks before the exam, the agent can do most of the synthesis work for you.

*"For the linear algebra final, pull every lecture summary, every problem-set learning note, every office-hours note, and every reading summary. Group them by topic. Flag what was emphasized for the exam."* You get a study guide grounded in your own notes — not a generic summary, not someone else's.

For practice problems specifically, the agent can extract them: *"From the problem sets and lecture transcripts, list every problem I marked as confusing. Suggest a study order."* Comes back with the prioritized list.

For the broader study-prep shape, the [Exam Prep and Competitive Study: A Vault That Actually Holds Your Material](/guides/personal-life/exam-prep-competitive-study/) approach extends this for high-stakes exams (bar exam, MCAT, professional certifications) — same vault mechanic, more intensive.

## A starter shape for the start of the term

A note on papers and essays before the starter list: a *Writing* sub-page per assignment holds the question, the outline, the sources, and the working draft. The agent can help with synthesis: *"For my paper on [topic], pull every relevant note from this term — readings, lectures, study group conversations — and outline how they connect to the thesis."* For the actual writing, the agent works inside the vault: *"Take the outline, draft the first section in my voice, grounded in the source notes I marked."* You edit the draft; the citations are traceable to specific pages in your vault. (For the broader writing-grounded-in-your-own-material shape, [How to Draft Emails, Proposals, and Newsletters Inside Your Notes App](/guides/creatives-content/draft-emails-proposals-in-notes/) covers it.)

If you're moving from notebooks and scattered Google Docs to a vault, this is what we'd suggest:

- **[Term name]** — top-level parent
- **[Course]** — one child per course
  - **Lectures** — one child per week, dated entries within
  - **Readings** — one child per assigned text or paper
  - **Problem sets** — one child per set
  - **Study group** — running notes per session
  - **Office hours** — running notes per visit
  - **Writing** — papers and essays
  - **Exam prep** — empty until midterms approach

That's it. The vault grows the way the term grows. The agent handles the synthesis at exam time, so you're not rebuilding the term from a stack of half-organized PDFs at 2 a.m. the week before finals.

The point isn't to turn studying into a project-management exercise. It's that the small amount of structure means the lectures stay searchable, the office-hours conversations don't evaporate, and the study sessions compound instead of restarting each week. (For pricing tiers, see [/pricing/](/pricing/) — most students do fine on the free tier through the term.)

[Try Docapybara free](/accounts/signup/) — start with the *Lectures* page for one course, record the next lecture, and the next time you go to review, the material will be in one place.