Turn Lecture Videos Into AI Study Guides: A Student Workflow
Convert 10+ hours of streaming lectures into searchable, AI-ready transcripts in 30 minutes
By the Mac Transcription Guide Team
Tested with real CS coursework across multiple universities
Table of Contents
The Problem
11 PM. Exam tomorrow. Three hours of lecture video between you and the one explanation you need.
The fix: Convert videos to searchable text. Let AI generate study materials from your actual course content.
⚡ Quick Start Option
Just need searchable transcripts? Complete Steps 1-2 only (15 minutes setup):
- Download lecture videos with Qooly
- Transcribe with Scriber Pro
Steps 3-5 unlock AI-powered study materials (optional but powerful).
What This Unlocks
Instant Search
Find any concept across an entire semester in seconds.
AI Analysis
Generate quizzes, outlines, and concept maps from your lectures.
Offline Access
No streaming hiccups before exams. Everything works offline.
Cross-Lecture Synthesis
AI connects concepts across different weeks automatically.
First Things First: The Terminal (It's Not Scary)
If you've never used a terminal before, here's the truth: it's just a text-based way to talk to your computer. That's it.
Instead of clicking on folders and files, you type commands. Why? Because for what we're doing—processing transcripts, organizing materials, running AI agents—it's actually faster and more powerful than clicking around.
What You'll Actually Do
Open Terminal (Mac: CMD+Space, type "Terminal")
Type a few commands
Talk to Claude
That's genuinely it. You're not hacking the mainframe. You're organizing study materials.
Getting Claude Code
Claude Code is Anthropic's official CLI tool. You need a Claude Pro membership ($20/month) to use it.
Installation
Open your terminal and run this:
$ npm install -g @anthropic-ai/claude-code
Then authenticate with your Claude account:
$ claude auth
It'll open a browser window. Log in. Done.
Your First Conversation
Navigate to your course folder and start Claude:
$ cd "/Users/yourname/Documents/FALL 2025/CS_321" $ claude
Claude starts. You can talk to it:
You: Hello! Can you see the files in this folder? Claude: Yes! I can see your course structure. You have: - Module 1 with 3 lecture transcripts - Module 2 with 2 lecture transcripts - A .claude/agents folder What would you like to work on?
That's it. You just talked to an AI that can read, analyze, and synthesize all your course materials.
The Real Magic: .md Files (Not Prompts)
Here's what most people get wrong: They think the prompt is the magic. It's not.
The magic is in the .md files you create. These are plain text files that contain your course materials—transcripts, notes, assignments, readings—and most importantly, links to other related files.
How It Works
You create markdown files that reference other files in your system using absolute paths:
# Introduction to Daoism Part 2 Focus on the figure of the sage in Zhuangzi's philosophy. ## Required Videos - [Professor's Introduction Video] '/Users/yourname/Documents/PHL_371/Module_2/Required_Video_transcription.txt' ## Required Readings - [In the World of Men, pages 49-62] '/Users/yourname/Documents/PHL_371/Module_2/In_the_World_of_Men_49-62.txt' - [The Sign of Virtue Complete, pages 63-72] '/Users/yourname/Documents/PHL_371/Module_2/Sign_of_Virtue_Complete_63-72.txt'
Why this matters: Notice how it links to the transcription of the professor's lecture video. When you ask Claude a question, it follows these links—reading the video transcription, the chapter readings, everything—then synthesizes an answer using your professor's actual words.
Example Query
Student: "The professor mentioned chapter 3 in the lecture video. What specific sections did he discuss?"
What Claude does:
- Reads Module Overview.md
- Follows link to lecture video transcription
- Searches transcription for "chapter 3" mentions
- Cross-references with actual reading file
- Tells you exactly which pages the professor covered
This is why the folder structure from earlier matters. This is why file naming matters. This is why linking matters. You're not just organizing files—you're building a knowledge graph that AI can navigate.
📚 Full Example: Real Philosophy Module
Here's a complete module file. Notice the links to lecture video transcriptions (created by Scriber Pro) and readings:
# Daoism Module 2 More Zhuangzi readings - builds on Module 1 concepts. Main focus: the concept of the sage and Zhuangzi's philosophy. ## Key Concepts - What is a sage in Daoist philosophy? - Zhuangzi's critique of society - How to live according to the Dao ## Required Videos Watch the intro video first (2 minutes): - [Professor's Video Introduction] '/Users/yourname/Documents/PHL_371/Module_2/Required_Video_transcription.txt' ## Required Readings Zhuangzi chapters (Watson translation): - [In the World of Men, pages 49-62] '/Users/yourname/Documents/PHL_371/Module_2/In_the_World_of_Men_49-62.txt' - [The Sign of Virtue Complete, pages 63-72] '/Users/yourname/Documents/PHL_371/Module_2/Sign_of_Virtue_Complete_63-72.txt' - [The Great and Venerable Teacher, pages 73-88] '/Users/yourname/Documents/PHL_371/Module_2/Great_Venerable_Teacher_73-88.txt' - [Fit for Emperors and Kings, pages 89-96] '/Users/yourname/Documents/PHL_371/Module_2/Fit_for_Emperors_Kings_89-96.txt' ## Discussion Assignment See instructions file after finishing readings: - [Module 2 - Discussion Instructions] '/Users/yourname/Documents/PHL_371/Module_2/Discussion_Instructions.md'
What You Can Ask:
"The professor mentioned Zhuangzi's view on death in the lecture—what exactly did he say and how does it connect to the reading?"
Claude follows the link to the lecture video transcription, finds the relevant section, cross-references with the reading file, and synthesizes an answer using your professor's actual words.
The key: Each .md file is an index pointing to your actual materials. Claude reads it, follows links, and navigates your file system like a knowledge graph.
💡 What's Actually Happening Here
The primary work you're doing isn't spending 6 hours passively watching lectures filled with 75% filler content.
The primary work is creating your .md files—deciding which materials connect, how concepts relate, what questions to ask. That's the actual intellectual labor.
Don't feel bad about this.
There's deep educational value in data collecting (organizing, structuring, mapping relationships) just as much as in data cramming (consuming content). When you build a knowledge graph by linking transcriptions to readings to discussions, you're doing the cognitive work of understanding how ideas connect—which is the whole point of education.
Read more: The Educational Value of Data Collecting
The Five-Step Workflow
1 Download the Videos
Most universities stream lectures in a format you can watch but not download (M3U8 files). Qooly is a free Chrome extension that captures them as regular MP4 files you can keep.
How it works: Install the extension, navigate to your lecture page, hit play, and Qooly grabs it. You can queue up a whole week's lectures and let them download overnight.
File naming: CS372_Lecture05_Network_Metrics.mp4 makes everything else easier.
2 Transcribe with Scriber Pro
This is the critical step. Scriber Pro turns hours of video into searchable text. Without this, none of the AI stuff works.
See How Easy It Is (30 seconds)
Watch a real transcription in action. Drag, drop, done.
Offline Processing
Everything stays on your Mac. No uploading course materials to random servers.
Batch Operations
Queue up 30+ hours, run it overnight. Wake up to searchable transcripts.
One-Time Cost
$3.99 total. Not per file. Not per month. Buy it once, use it forever.
The process: Open Scriber Pro, drag in your lecture videos, select TXT as output format, hit transcribe. A 90-minute lecture processes in about 6-8 minutes on an M1/M2 Mac.
Output: CS372_Lecture05_Network_Metrics.txt appears next to the MP4, ready for AI processing.
🔑 Why Scriber Pro is the Bridge to Everything Else
Here's the fundamental problem: Claude can't read video. No AI can. Video is just pixels and audio waves. AI needs text.
Scriber Pro is the translator. It takes your lecture videos—whatever format they're in (MP4, MOV, M4V, whatever)—and turns the spoken words into plain text files that Claude can actually read.
The transformation:
(Claude: ❌ Can't read)
(Claude: ✅ Can read)
Without Scriber Pro, this entire workflow is impossible. You can't link to video files in your .md files and expect Claude to understand them. You can't search video. You can't ask AI to synthesize concepts from video. The video has to become text first.
That's why this step isn't optional. Scriber Pro is the bottleneck—in a good way. Everything downstream depends on it converting your lectures from unsearchable video into AI-readable text.
3 Organize Systematically
Folder structure determines how AI agents understand your courses. Do it right once.
Here's how to actually do it:
📖 Command Cheat Sheet (Don't Worry, It's Easy)
cd
Change Directory — Navigate into a folder. Think of it like double-clicking a folder, but by typing its name.
mkdir
Make Directory — Create a new folder. Same as right-click → New Folder.
mv
Move — Move a file from one place to another. Like dragging and dropping.
That's it. Three commands. You're not writing code—you're telling your computer where to put stuff.
# Go to your Documents folder $ cd "~/Documents/FALL 2025" ← Navigate to your semester folder # Create a new folder for this course $ mkdir "CS_321" ← Make a folder called "CS_321" # Go into that new folder $ cd CS_321 ← Open the CS_321 folder # Create folders for each module $ mkdir "Module 1" "Module 2" "Module 3" ← Make 3 folders at once # Move transcript files into Module 1 $ mv CS321_Lecture01*.txt "Module 1/" ← Move Lecture 1 files $ mv CS321_Lecture02*.txt "Module 1/" ← Move Lecture 2 files
What you end up with:
FALL 2025/ └── CS_321/ ├── .claude/agents/ │ └── document-quiz-ingestor.md ├── Module 1/ │ ├── CS321_Lecture01_Introduction.txt │ ├── CS321_Lecture02_Math_Prelim.txt │ └── CS321_Lecture03_DFA_Basics.txt ├── Module 2/ │ ├── CS321_Lecture04_NFA_Intro.txt │ └── CS321_Lecture05_NFA_to_DFA.txt └── Module 3/ ├── CS321_Lecture06_Regex.txt └── CS321_Lecture07_Grammars.txt
Why this works: AI agents can process by module, course, or entire semester. The hierarchy mirrors how concepts build.
4 Configure AI Agents Advanced
AI agents analyze transcripts and generate study materials on demand. No more manual note compilation.
What agents do:
- Search all lectures to find every mention of a concept
- Generate practice quizzes from lecture content
- Create study outlines showing concept connections
- Build concept maps of topic relationships
- Answer questions by synthesizing multiple lectures
Example agent configuration (real CS student workflow):
KNOWLEDGE STRUCTURING: - Organize information into logical knowledge clusters - Create mental maps of concept relationships - Identify cause-and-effect, comparisons, contrasts QUIZ RECALL READINESS: - Answer questions from basic recall to complex analysis - Provide accurate, complete answers with detail - Cite specific lectures or sections when relevant CROSS-DOCUMENT SYNTHESIS: - Track how concepts evolve across lectures - Identify when lectures discuss related topics - Build comprehensive understanding from distributed info
Real use case: "Explain all instances where the professor discussed NP-completeness and connect to reduction techniques." The agent searches transcripts from Lectures 3, 7, 9, and 14, then synthesizes a comprehensive answer.
5 Generate Study Materials Advanced
With organized transcripts and configured agents, create custom study materials instantly.
Practice Exams
"Generate 10 questions on TCP/IP with graduated difficulty"
Study Outlines
Hierarchical summaries of entire modules
Concept Maps
Visual relationships between topics
Reference Sheets
Formula compilations for open-book exams
The difference: Materials come from your actual lectures. The terminology, examples, and emphasis match what your professor teaches.
Start to Finish: Implementation
What You Need
- Mac computer (for Scriber Pro)
- Chrome browser (for Qooly)
- 2-3 GB storage per course
- Claude Code (optional, for AI features)
The Setup Sequence
- Install Qooly: Get it from qooly.com
- Get Scriber Pro: Mac App Store ($3.99 one-time)
- Download lectures: Use Qooly to capture streaming videos
- Transcribe: Batch process all videos with Scriber Pro
- Organize: Create folder structure mirroring course hierarchy
- Configure agents: Set up AI processing (optional but powerful)
Start small: Test with one course. Refine naming and structure before scaling to full semester.
The Fine Print
Academic Ethics
Personal educational use only. Download only content you have access to. Don't redistribute lectures or transcripts. Check your institution's acceptable use policies.
Technical Reality
- Transcription accuracy: High quality but not perfect. Verify critical details (equations, heavy accents) against original video
- Visual content: Diagrams don't transcribe. Keep original videos for reference
- Platform compatibility: Works with most streaming platforms, not all proprietary systems
Why This Approach Works
Time Liberation
Typical course: 30+ hours of lecture video. Traditional review means scrubbing through hours of content to find specific information.
With this workflow: Text search finds anything instantly. AI synthesis generates study materials in minutes.
Setup investment: 4-6 hours per semester. Pays back in efficiency every single week.
Scriber Pro is Non-Negotiable
I'm going to be blunt: without Scriber Pro, this whole thing falls apart. Video stays video. AI can't read it. You're stuck scrubbing through hours of footage looking for one explanation.
Why Scriber Pro specifically:
- Scale: Process an entire semester (30+ hours) for $3.99. Cloud services would cost you $450+ at $0.25/minute.
- Privacy: Everything happens offline on your Mac. No uploading lectures to third-party servers (which often violates university policies).
- Accuracy: It actually handles technical terminology. "Dijkstra's algorithm" doesn't become "deck straw's algorithm."
- Batch processing: Queue everything up, let it run overnight. Done.
Other approaches either cost too much, violate privacy policies, or take forever. Scriber Pro solves all three problems for less than a cup of coffee.
Start Building Your Workflow
Get all three tools: Qooly to download lectures, Scriber Pro to transcribe them, and Claude Code for AI-powered learning.
Free • Chrome/Edge • Download lectures
$3.99 one-time • Unlimited transcriptions
$20/month Pro • Free trial available
Quick Recap
- Download: Capture streaming lectures with Qooly
- Transcribe: Convert to searchable text with Scriber Pro (offline, batch)
- Organize: Structure folders to mirror course hierarchy
- Configure: Set up AI agents to analyze materials
- Generate: Create custom study materials on demand
Result: Instant search across semesters. AI-powered study tools. Automated material generation from your actual course content.