YOUR PDF KNOWLEDGE SYSTEM — MASTER GUIDE Organise · Navigate · Activate

YOUR PDF KNOWLEDGE SYSTEM

A Complete Guide to Managing, Navigating & Activating Your
Research Library

How to use every PDF you've created — when to open which one, how to connect them, and how to turn
pages into action.

§0 Your PDF Library — Complete Catalogue

# PDF Title Type Colour Theme

Independent Advanced
01 Research Roadmap Research Strategy Purple / Cyan
(AI/Math/CS/Physics)

Engineering Research
02 Roadmap (this companion Research Strategy Navy / Multi
doc's pair)

PDF Management Guide

03 Meta / Operations Indigo / Rainbow
(this document)

Any other PDFs you've

... Various Various
created in prior sessions

Why This Guide Exists: When you have multiple dense reference PDFs, the biggest risk is cognitive
overload — opening everything, reading nothing deeply, acting on nothing. This guide solves that. It gives
you a system: a folder structure, a daily ritual, a decision tree for 'which PDF do I open right now,' and a
method for extracting action from each document.

The Core Principle: A PDF is a tool, not a trophy. Each one should be opened with a specific question,
produce a specific action, and then be closed. This guide teaches you to use them that way.

Your Personal Research & Knowledge Library Page 1

YOUR PDF KNOWLEDGE SYSTEM — MASTER GUIDE Organise · Navigate · Activate
SECTION 1
Folder Architecture
How to store, name, and version your PDFs

The Master Folder System

Create this exact folder structure on your computer, Google Drive, and/or Notion:

■ RESEARCH_LIBRARY/ ■■■ ■ 00_META/ ■ ■■■ PDF_Management_Guide.pdf ← this document ■ ■■■

Master_Index.xlsx ← spreadsheet: all PDFs, tags, last-opened, action items ■ ■■■
Weekly_Review_Template.md ← your Sunday ritual checklist ■ ■■■ ■ 01_RESEARCH_ROADMAPS/ ■
■■■ AI_Math_CS_Physics_Roadmap.pdf ← research directions, 3-paper plan ■ ■■■
Engineering_Roadmap.pdf ← 4 projects across 4 disciplines ■ ■■■ _NOTES/ ← your
handwritten/typed notes on each roadmap ■ ■■■ ■ 02_ACTIVE_PROJECTS/ ■ ■■■
Project_Mech_TopologyOpt/ ■ ■■■ Project_Aero_TPS/ ■ ■■■ Project_Elec_SNN/ ■ ■■■
Project_Chem_Reactor/ ■ ■■■ ■ 03_PAPERS_AND_LITERATURE/ ■ ■■■ _To_Read/ ■ ■■■
_Read_Annotated/ ■ ■■■ _Cited_In_My_Work/ ■ ■■■ ■ 04_DRAFTS_AND_WRITING/ ■ ■■■
Paper1_Draft_v1.pdf ■ ■■■ Paper1_Draft_v2.pdf ■ ■■■ ■ 05_ARCHIVE/ ■■■ (older versions,
superseded PDFs)

Naming Convention Rules

Rule Format Example

2025-07-01_AI_Research_Roadmap.p
Date-prefix all files YYYY-MM-DD_Title.pdf
df

Use underscores, not spaces No_Spaces_Ever.pdf Engineering_Roadmap_v2.pdf

Version suffix for drafts _v1, _v2, _FINAL Paper1_Draft_v3_FINAL.pdf

Tag type in filename [ROADMAP], [PAPER], [NOTES] [ROADMAP]_Engineering_2025.pdf

Never delete — archive Move to 05_ARCHIVE/ 01_OLD_Engineering_Roadmap.pdf

The Master Index Spreadsheet

Create a file called Master_Index.xlsx with these columns for every PDF in your library:

• Filename — exact filename

• Category — Roadmap / Paper / Notes / Literature / Tool

• Purpose — one sentence: what is this PDF for?

• Last Opened — date (update manually or with Notions/Obsidian automation)

• Status — Active / Reference / Archive

• Key Action Items — 1–3 specific things this PDF tells you to do

• Linked PDFs — which other PDFs reference or connect to this one

Review this index every Sunday. If a PDF hasn't been opened in 4 weeks and has no upcoming
action item, move it to Archive.

Your Personal Research & Knowledge Library Page 2

YOUR PDF KNOWLEDGE SYSTEM — MASTER GUIDE Organise · Navigate · Activate
SECTION 2
When to Use Which PDF
Decision tree and daily trigger guide

The Trigger-Based Decision System

The most common mistake with reference PDFs is opening them randomly and reading passively. Instead, use a
trigger-based system: every PDF
AI/Math/CS/Physics has specific 'trigger conditions' — situations where that PDF is the right tool.
Roadmap
■
■
Engineering
Here's the complete
This
You'rePDF
Roadmap
trigger map: Guide
Management
■ deciding what to research next week, or feeling lost about research direction. Open this to re-anchor on your 3-paper plan and m
You're starting a new engineering project, need a methods checklist for your discipline, or want to see how your work connects to othe
You feel overwhelmed by too many PDFs, or you're setting up your system for the first time, or doing your Sunday weekly review.

The Daily Decision Tree

START HERE → What is your goal right now? ■■ "I need to decide what to work on today" ■
■■ Open: Research Roadmap → check monthly milestone → pick top task ■ ■■ "I'm starting
a new experiment or implementation" ■ ■■ Open: Relevant Engineering or Research Roadmap
section ■ → copy the methodology steps → open your lab notebook ■ ■■ "I need to write
or structure a paper" ■ ■■ Open: The roadmap's 'Publication Plan' section for that paper
■ → check target venue + scope → open Overleaf ■ ■■ "I feel overwhelmed / don't know
where to start" ■ ■■ Open: This PDF (Management Guide) → Section 4 (Weekly Ritual) ■ →
do the 15-minute reset exercise ■ ■■ "I need to pick a specific tool, library, or
method" ■■ Open: Relevant roadmap → Tools & Resources section → pick tool → open its
documentation

Context-Specific Usage Table

Your Situation Open This PDF Go To Section Do This

Monday morning — Pick 1 milestone to complete

Research Roadmap Publication Timeline
planning week this week

Starting FEM / CFD Mechanical or Aerospace Copy methodology steps to

Engineering Roadmap
simulation section lab notebook

Choosing a journal to submit Check target venues table,

Either Roadmap Publication Strategy
to verify deadline

Stuck on a research Interdisciplinary Landscape Find adjacent field that may

Research Roadmap
problem §3 offer a solution

Follow the tool install +

Setting up a new tool/library Engineering Roadmap §5 Universal Infrastructure
validation steps

§4 Research Direction for Use Problem Statement as

Writing paper introduction Research Roadmap
that paper intro scaffolding

Executive Summary + Extract 3 core claims, build

Preparing a conference talk Both Roadmaps
Research Direction slides around them

Feeling directionless / Run the Sunday Reset

This Guide Section 4: Weekly Ritual
overwhelmed protocol

Applying to Anthropic / BAIR Follow institution-specific

Research Roadmap §7 Strategic Advice
/ Jane St. strategy

Your Personal Research & Knowledge Library Page 3

YOUR PDF KNOWLEDGE SYSTEM — MASTER GUIDE Organise · Navigate · Activate
SECTION 3
Active Reading & Annotation System
How to read a dense PDF so it produces action, not just highlights

The 3-Pass Reading Method

For any substantial PDF (roadmap, technical guide, research paper), use the 3-pass method. Never read
start-to-finish on firstPass
The Skim pass (5–8
— that's how dense documents become overwhelming.
minutes)
1
Read ONLY: title, executive summary, section headers, tables, and call-out boxes. Goal: understand the
structure. Do NOT take
The Structure notes
Pass yet. minutes)
(15–20
2
Read the introduction to each section + the first and last paragraph of each sub-section. Mark sections
relevant
ThetoDeep
your Pass
current(variable
project. Write
— only1 question
relevantpersections)
section in the margin.
3
Read only the sections you marked in Pass 2. For each paragraph, write 1 action item or 1 connection to
your existing work. Stop when you have 5 action items total.

Annotation Colour Code (PDF Reader)

Colour Meaning When to Use

Yellow highlight Important fact or concept Information you'll need to recall later

Blue highlight Actionable step or method Something you will actually implement

Orange highlight Connection to another PDF Note which other PDF this connects to

Something you think is wrong or

Red highlight / underline Disagreement or gap incomplete — potential research
contribution

Green highlight Citation to track down Papers you need to read

Your original thought triggered by this

Comment / sticky note Your own insight
content — potential paper idea

After Reading: The 5-Output Rule

Rule: After finishing a reading session with any PDF, you must produce exactly 5 outputs before closing it:

1. 3 Action Items — specific, date-stamped tasks (e.g., 'Install FEniCSx by Friday')

2. 1 Connection — one explicit link to another PDF in your library (e.g., 'This methodology section
connects to Section 4.1 of the Research Roadmap')

3. 1 Open Question — something the PDF raised that it didn't answer — this is a potential research gap

Write these 5 outputs in your Master_Index.xlsx under 'Key Action Items' for that PDF before you
close it. If you can't produce 5 outputs, you haven't read deeply enough — go back to Pass 3.

Your Personal Research & Knowledge Library Page 4

YOUR PDF KNOWLEDGE SYSTEM — MASTER GUIDE Organise · Navigate · Activate
SECTION 4
The Weekly Review Ritual
A 45-minute Sunday protocol to keep your PDF library alive and working

Why Weekly Review Matters

PDFs decay if not actively revisited. A roadmap you read once but never review becomes dead weight.
The weekly ritual is what transforms a PDF library from a collection of intentions into a living research
system. Block 45 minutes every Sunday evening.

Capture (5 min)
The
1 45-Minute Sunday Protocol
Collect all loose notes, highlights, Post-its, browser tabs related to your PDFs. Dump everything into a
single 'Inbox' note.
Process Inbox (8 min)
2
For each item in Inbox: is it an action? → add to Master_Index action column. Is it a connection between
PDFs? → note
Review Active irrelevant? →
it. Is itRoadmaps delete.
(10 min)
3
Open your primary roadmap (Research or Engineering). Check: which milestone was I supposed to hit this
week? Did I? Master
Update If not, why? Reschedule
Index (7 min) or adjust.
4
Update 'Last Opened' dates, status changes (Active → Archive), and action item completion in
Master_Index.xlsx.
Identify Next Week's Anchor PDF (5 min)
5
Pick Cross-Link
ONE PDF that will be(5your
Check min)primary reference next week. Write it at the top of your Monday to-do list.
6
Look at your open questions from last week's reading. Does any other PDF in your library answer them?
Create an explicit
Archive Sweep link note.
(5 min)
7
Any PDF with no action items and last-opened >4 weeks ago: move to 05_ARCHIVE. Ruthless is good.

The 15-Minute 'Overwhelm Reset'

When to use: When you feel buried by too many PDFs, too many action items, or just can't decide what to
do.

Step 1 (3 min): Close everything. Open ONLY this PDF. Breathe.

Step 2 (3 min): Write down the ONE thing that, if completed today, would make the most progress on your
goals.

Step 3 (3 min): Identify which single PDF contains the information you need for that one thing.

Step 4 (6 min): Open ONLY that PDF. Do Pass 2 on the relevant section. Extract the 3 action items. Close
the PDF.

Result: You now have a clear task from a specific source. Everything else can wait.

Your Personal Research & Knowledge Library Page 5

YOUR PDF KNOWLEDGE SYSTEM — MASTER GUIDE Organise · Navigate · Activate
SECTION 5
Connecting Your PDFs — The Knowledge Graph
How your documents reinforce each other across disciplines

The PDF Connection Map

Your PDFs are not isolated documents — they form a network. Understanding the connections between them lets
you use insights from one to accelerate work in another. Here is the connection map for your current library:

Research Roadmap (AI/Math/CS/Physics) → Engineering Roadmap:

• The PINN aerothermal surrogate in Aerospace (Eng. Roadmap §4.2) IS the Physics-Informed Neural
Network discussed in Research Roadmap §4.1. The same architecture, different application domain.

• The GNN surrogate for FEM in Mechanical Engineering connects directly to Research Roadmap §4.1
(geometric deep learning on non-Euclidean domains — a graph is a mesh).

• KV-cache compression research (Research Roadmap §4.3) uses information theory techniques identical
to those in Chemical Engineering Bayesian optimisation (Engineering Roadmap §4.4 Track C).

Action: When you make progress on any one of these, immediately check if the same technique applies in
the connected domain. This is how interdisciplinary researchers generate 2x output from 1x effort.

Cross-PDF Methodology Reuse Table

Method / Technique Roadmap 1 (Research) Roadmap 2 (Engineering) Reuse Strategy

Mech: FEM surrogate; Same PyG code, different

Graph Neural Networks §4.1: NeRF scene rep
Chem: catalyst screening graph types

Bayesian Optimisation Chem: reactor operating Same BoTorch loop,

§4.3: KV-cache tuning
(BoTorch) conditions different objective function

Physics-Informed Neural Aero: aerothermal CFD Same architecture, different

§4.1: spatial intelligence
Nets surrogate governing PDE

Analogous circuit-level
Mechanistic interpretability §4.2: transformer circuits Elec: SNN neuron analysis analysis on different
substrate

Shannon entropy as
Chem: Bayesian acquisition
Information theory (entropy) §4.3: KV-cache pruning universal importance
function
measure

§4.2: logic-constrained Elec: async circuit timing Lean 4 proofs apply to both
Formal verification
training proofs software and hardware

Building Your Personal Knowledge Graph

Tool Option A — Obsidian (Recommended): Tool Option C — Zotero: Best for managing
Free, local markdown editor with a graph view. academic papers alongside your PDFs. Has a
Create one note per PDF. Link notes using 'Related' field per item. Integrates with Obsidian via
[[PDF_Name]] syntax. The graph view visualises the Zotero-Obsidian plugin.
your connections automatically. Add tags like
#methods, #robotics, #ML to each note.

Your Personal Research & Knowledge Library Page 6

YOUR PDF KNOWLEDGE SYSTEM — MASTER GUIDE Organise · Navigate · Activate

Tool Option B — Notion: Create a database table Minimum Viable Graph (no tools): A single A4
(= your Master Index). Each row is a PDF. Add a paper with each PDF as a node (circle) and
'Related PDFs' relation column. Notion's built-in hand-drawn arrows between connected PDFs.
graph is less visual but more structured. Tape it to your wall. Primitive but effective —
researchers at top labs often have exactly this.

Your Personal Research & Knowledge Library Page 7

YOUR PDF KNOWLEDGE SYSTEM — MASTER GUIDE Organise · Navigate · Activate
SECTION 6
Digital vs. Print Strategy
When to read on screen, when to print, and how to annotate effectively

The Golden Rule: Print for Strategy, Screen for Reference

Print These (Strategy Documents): Keep These Digital (Reference Documents):

• Research Roadmap — tape the timeline page to • Full PDF text — search is invaluable for dense
your wall reference material

• Engineering Roadmap cover + master timeline — • Literature papers — annotate with Zotero + PDF
above your desk reader

• This guide's Section 2 (Decision Tree) — next to • Draft papers — track changes, version control
your keyboard

• Your current milestone list from your active • Master Index — update it, it must be editable
roadmap

Why print strategy docs: Physical documents Why keep reference digital: Ctrl+F is faster than
activate spatial memory differently. When a flipping pages for finding a specific method,
roadmap is on your wall, your brain processes it equation, or tool. Also, digital annotations are
passively throughout the day. You'll notice searchable — you can find all your blue 'action
connections and remember milestones without item' highlights across multiple PDFs at once.
actively reading.

Recommended PDF Reader Setup

Platform Recommended Reader Key Feature to Enable Why

Sumatra PDF (for reading) + Keyboard shortcuts: G (go Sumatra is instant-open;

Windows
Okular (for annotation) to page), F3 (find) Okular has full annotation

Preview (built-in) + PDF Sidebar thumbnails always Preview is fast; PDF Expert
macOS
Expert (paid) visible has superior highlights

Enable annotation toolbar by Okular annotations are

Linux Okular or Evince
default exportable as text

Create a 'Research Library'

Apple Pencil annotation =
iPad GoodNotes 5 or Notability notebook, one section per
best of print + digital
PDF

Enable reading mode for Xodo handles large PDFs

Android Xodo PDF or Adobe Reader
long sessions without crashing

Zotero (as a library Enable 'Attached files' Keeps your PDFs organised
All platforms
manager) auto-rename and sync with metadata automatically

Keyboard Shortcuts That Save Hours

Universal PDF shortcuts (most readers):

• Ctrl+F / Cmd+F — Find text. USE THIS CONSTANTLY. Never page-flip to find something.

Your Personal Research & Knowledge Library Page 8

YOUR PDF KNOWLEDGE SYSTEM — MASTER GUIDE Organise · Navigate · Activate

• Ctrl+G / Cmd+G — Go to specific page number. Know your key sections' page numbers.

• Ctrl+Home/End — Jump to first/last page instantly.

• Space / Shift+Space — Scroll down/up by one screen.

• + / - — Zoom in/out. Use 125% zoom for dense technical text.

• Right/Left arrow — Next/previous page (works in most readers).

Obsidian shortcuts (if using as knowledge graph):

• Ctrl+O — Quick open any note. Ctrl+E — Toggle edit/preview. Ctrl+Shift+F — Search all notes.

Your Personal Research & Knowledge Library Page 9

YOUR PDF KNOWLEDGE SYSTEM — MASTER GUIDE Organise · Navigate · Activate
SECTION 7
Future-Proofing Your Library
What to do as you generate more PDFs, papers, and notes

When Your Library Grows Beyond 10 PDFs

The 10-PDF Threshold: Below 10 PDFs, a simple folder structure and weekly review is sufficient. Above
10, you need tags, search, and automated reminders. Here's the upgrade path:

10–20 PDFs: Add a 'Tags' column to Master_Index.xlsx. Use tags like: #active, #reference, #archive,
#methods, #project-mech, #project-aero, #paper1, #reading-queue.

20–50 PDFs: Migrate Master Index to Notion or Airtable. Enable full-text search across all your notes
about PDFs (Obsidian does this). Create filtered views: 'Show me all active PDFs with unfinished action
items.'

50+ PDFs: You now have a personal research library. Use Zotero for literature, Obsidian for your own
documents and notes, and a simple sync solution (Google Drive or Dropbox) for backup. Schedule a
monthly 'Library Audit' in addition to weekly review.

When to Commission a New PDF (vs. Add Notes to Existing)

Create a new PDF when: Add notes to existing PDF when:

• You're starting a fundamentally new research • You've made progress on an existing project
direction (annotate milestones as DONE)

• You need a structured roadmap for a new • You found a new paper that extends an existing
discipline or project research direction

• You want a reference document to share with • A methodology step needs an update or
collaborators correction

• The new content would double the length of an • You have a new tool that replaces a
existing PDF recommended tool

• You need a new timeline/milestone structure • The new insight is a refinement, not a new
direction

Backup & Sync Strategy

Priority Action Frequency Tool

Backup entire
Google Drive auto-sync or
Critical RESEARCH_LIBRARY Daily (automated)
Dropbox
folder

Version control for code +

High Every commit Git + GitHub
data referenced in PDFs

Export Obsidian vault / Obsidian sync or Notion

Medium Weekly (Sunday ritual)
Notion database export

Manual move to
Low Archive old PDF versions Monthly
05_ARCHIVE

Your Personal Research & Knowledge Library Page 10

YOUR PDF KNOWLEDGE SYSTEM — MASTER GUIDE Organise · Navigate · Activate

The Single Most Important Habit

Every PDF you create or read should produce at least one dated action item before you close it. It
doesn't matter how sophisticated your folder structure is, how elegant your knowledge graph looks, or how
many colours you use for annotations. If a PDF session doesn't end with a specific action written in your
task manager, the session was consumption, not work. Research is not reading. Research is doing
something that only you have done before — and your PDFs are the maps that tell you what to do next.
Use them as maps. Not as proof that you studied.

Your Personal Research & Knowledge Library Page 11