OpenLawNotes

Automatically transcribe lecture recordings and summarize readings into clean, structured notes — powered by AssemblyAI and Google Gemini.

---

⚠️ Recording Permissions

This tool processes audio recordings of lectures. Before recording any class session:

• Always get explicit permission from your professor before recording
• Check your school's academic policies on lecture recording
• Many professors prohibit recording — never record without consent
• Some jurisdictions have two-party or all-party consent laws for audio recording

This project takes no responsibility for recordings made without proper authorization. By using this tool, you confirm that all recordings were made with appropriate permission.

⚠️ University AI Policy

This tool uses AI services (AssemblyAI and Google Gemini) to process lecture content and generate notes. Before using this tool:

• Always check your university's AI use policy before using AI-generated content for academic work
• Many institutions restrict or prohibit the use of AI tools for coursework — ensure your use complies with your school's rules
• AI-generated notes are a study aid; review them critically and supplement with your own understanding
• Some courses or exams may prohibit the use of AI-assisted materials entirely

This project takes no responsibility for any violations of your institution's academic integrity or AI use policies. By using this tool, you confirm that your use complies with all applicable university policies.

---

What It Does

• Transcribes lecture audio (.m4a files) using AssemblyAI — with speaker identification so you can tell professor from student
• Saves the full transcript so you can verify the AI's notes against the actual words spoken
• Generates structured lecture notes from transcripts using Google Gemini
• Summarizes readings (PDF, DOCX, TXT, MD) into concise study notes
• Copies all new notes to a single folder for quick one-stop access after every run

---

How It Works

1. Drop your lecture recording into the lecture-input/ folder inside your class folder
2. Drop your reading (PDF, DOCX, TXT, or MD) into the reading-input/ folder
3. Run python main.py
4. Pick up your notes:
   • lecture-output/ — lecture notes for that class
   • reading-output/ — reading summaries for that class
   • new-outputs-safe-delete/ (in the project folder) — every note generated this run, all in one place for easy grabbing

After processing, your original files are moved (not deleted) into lecture-processed/ and reading-processed/. The full lecture transcript is saved to lecture-processed/txt/ — open it any time to cross-check the AI's notes against what was actually said.

The new-outputs-safe-delete/ folder is a convenience copy of the latest outputs. Once you've grabbed what you need, you can safely delete its contents — the real notes always live in the class folders.

---

Accepted audio formats

The tool accepts M4A files only (*.m4a). This is the default format for the most common recording apps:

App	Platform	Default format
Voice Memos	iPhone / Mac	M4A ✓
Voice Recorder / Sound Recorder	Windows 10/11	M4A ✓
Zoom (audio-only local recording)	Windows / Mac	M4A ✓

If your recorder saves a different format (MP3, WAV, etc.), convert it to M4A first. Free tools like Audacity or CloudConvert can do this in a few clicks.

---

Quick Start (< 5 minutes)

Step 1 — Install Python

Download Python 3.10 or newer from python.org. During installation, check "Add Python to PATH".

Verify it worked — you'll do this in a terminal.

What is a terminal? It's a plain text window where you type commands and press Enter to run them.

• Windows: Press the Start button, type Command Prompt, and press Enter.
• Mac: Press Command + Space, type Terminal, and press Enter.

Type the command for your system into the terminal and press Enter:

Windows:

python --version

Mac:

python3 --version

You should see something like Python 3.12.0. If you get an error instead, Python was not installed correctly — re-run the installer and make sure to check "Add Python to PATH" before clicking Install.

Step 2 — Install FFmpeg

FFmpeg converts your .m4a recordings to the format AssemblyAI expects. You only need to do this once.

Windows — Install with Chocolatey. Once Chocolatey is installed, open an Administrator PowerShell (press Start, type PowerShell, right-click it, and choose "Run as administrator"), then run:

choco install ffmpeg

Mac — Install with Homebrew. Once Homebrew is installed, open Terminal and run:

brew install ffmpeg

Verify it worked — in your terminal (Command Prompt on Windows, Terminal on Mac), run:

ffmpeg -version

You should see a version number printed out. If you get an error, restart your terminal and try the install step again.

Step 3 — Install OpenLawNotes

Download the project. Click the green Code button on this GitHub page and choose Download ZIP. Unzip it somewhere easy to find — like your Desktop or Documents folder. Note the full path to the folder (you'll need it in a moment).

Open a terminal and navigate to the project folder.

What does "navigate" mean? You use the cd command (short for "change directory") to tell the terminal which folder to work in. You only need to do this once per session.

Windows — open Command Prompt and run (replace the path with wherever you unzipped the project):

cd C:\Users\YourName\Desktop\open-law-notes

Mac — open Terminal and run:

cd /Users/YourName/Desktop/open-law-notes

Tip: The easiest way to get the right path is to open the folder in File Explorer (Windows) or Finder (Mac), then copy the address from the address bar. On Windows you can also type cmd directly into the File Explorer address bar to open a terminal already in that folder.

Create a virtual environment. This keeps OpenLawNotes's packages separate from anything else on your computer. Run this once, from inside the project folder:

Windows:

python -m venv venv

Mac:

python3 -m venv venv

Activate the virtual environment. You must do this every time you open a new terminal before running the tool:

Windows:

venv\Scripts\activate

Mac:

source venv/bin/activate

Your terminal prompt will change to show (venv) at the start — that means it worked.

Every time you want to run OpenLawNotes: open a terminal, cd to the project folder, then activate the venv. The QUICKSTART.md file in this project has all three commands in one place for easy copy-paste.

Install the required packages (only needed once, after activating the venv):

Windows:

pip install -r requirements.txt

Mac:

pip3 install -r requirements.txt

Step 4 — Get API Keys

You need two free API keys:

AssemblyAI (for transcription)

1. Sign up at assemblyai.com
2. Copy your API key from the dashboard

Google Gemini (for note generation)

1. Go to aistudio.google.com
2. Click "Get API key" and copy it

Step 5 — Configure

Copy the example env file:

Windows

copy .env.example .env

Mac

cp .env.example .env

Open .env and paste your keys:

GEMINI_API_KEY=paste-your-gemini-key-here
ASSEMBLYAI_API_KEY=paste-your-assemblyai-key-here

Copy the example config file:

Windows

copy config.py.example config.py

Mac

cp config.py.example config.py

Open config.py and set your PARENT_FOLDER — the folder that contains all your class folders:

PARENT_FOLDER = "C:/Users/YourName/Documents/Law school"

How to find this path:

• Windows — Right-click your law school folder in File Explorer and click Copy address. This gives you something like C:\Users\YourName\Documents\Law school. When pasting into config.py, replace each \ with / (forward slash) — for example: "C:/Users/YourName/Documents/Law school". Forward slashes always work on Windows in Python.
• Mac — Open Finder and navigate to your law school folder. Right-click (or Ctrl-click) it while holding the Option key, then click "Copy [folder name] as Pathname".

Then list your classes:

CLASSES = {
    "Contracts": None,
    "Civ Pro":   None,
}

Each name must exactly match the folder name on your computer. That's it. Run the pipeline:

python main.py

You should see step-by-step progress logged to your terminal.

---

File Structure

Each class folder gets an open-law-notes/ subfolder automatically created on the first run. Drop your files in the input folders before running.

Your Class Folder/
└── open-law-notes/
    ├── lecture-input/           ← DROP .m4a recordings here before running
    ├── lecture-output/          ← Lecture notes appear here after running ✓
    ├── lecture-processed/
    │   ├── audio/               ← Original .m4a moved here after processing (preserved)
    │   └── txt/                 ← Full transcripts saved here — open to verify AI notes
    ├── reading-input/           ← DROP PDFs, DOCX, TXT, MD files here before running
    ├── reading-output/          ← Reading summaries appear here after running ✓
    └── reading-processed/       ← Original reading files moved here after processing (preserved)

Files are moved (not deleted) into the processed folders after each run, so your originals are always safe.

Every output is also copied to new-outputs-safe-delete/ in the project root — a single folder with everything generated on the latest run. Grab what you need, then delete the contents whenever you like. The real notes live in the class folders above.

---

Updating for a New Semester

Open config.py and update the CLASSES dict to your new class names. That's it — the folder structure is auto-created fresh for each new entry.

CLASSES = {
    "Contracts": None,
    "Civ Pro":   None,
}

---

Running the Pipeline

Full run (transcribe audio + generate all notes):

python main.py

Reading-only mode (skip audio, only process reading files):

python main.py --read-only

---

Estimated Costs

Both APIs have free tiers, but charges apply once you exceed them. Rough estimates:

What	Estimated cost
Lecture (transcription + notes)	~$0.30 per hour of audio
Reading summary	~$0.15 per reading assignment

Costs come from AssemblyAI (transcription) and Google Gemini (note generation). Both offer free monthly credits that will cover light usage before any charges kick in.

---

Configuration Reference

All settings live in config.py (your local copy of config.py.example).

Setting	Default	Description
PARENT_FOLDER	(your path)	Path to the folder containing all your class folders
CLASSES	(your classes)	Dict of class names → Drive folder ID (or None)
ENABLE_GOOGLE_DRIVE	True	Set False to skip Google Drive download (Step 0)
ENABLE_DIARIZATION	True	Identify speakers in transcripts
MAX_SPEAKERS	None	Speaker count hint for diarization (None = auto)
GEMINI_MODEL	gemini-3.1-pro-preview	Gemini model for note generation
MAX_AUDIO_WORKERS	2	Parallel audio upload threads
MAX_LLM_WORKERS	5	Parallel LLM processing threads
LLM_BASE	open-law-notes	Name of the auto-created subfolder in each class folder
NEW_OUTPUTS_DIR	new-outputs-safe-delete	Local folder where copies of all outputs are placed

---

Google Drive Setup (Optional)

If ENABLE_GOOGLE_DRIVE = True, the pipeline will automatically download new .m4a audio files from your Google Drive before processing. Each class can have its own Drive folder.

Why use this? If you record lectures on a phone or another device, uploading the recording to Google Drive and letting the pipeline download it automatically is an easy way to move files over without manually copying them to your computer.

To set up:

1. Go to Google Cloud Console and create a project
2. Enable the Google Drive API
3. Create OAuth 2.0 credentials (Desktop app), download the JSON, and save it as credentials.json in the project root
4. For each class you want to sync, find its Drive folder ID:
   • Open the class folder in Google Drive in your browser
   • Look at the URL — it will look like:

     https://drive.google.com/drive/folders/1aBcDeFgHiJkLmNoPqRsTuV
     

   • The long string of letters and numbers at the end is the folder ID. Copy it.
5. Paste each ID into config.py next to the class name:

   CLASSES = {
       "Contracts": "1aBcDeFgHiJkLmNoPqRsTuV",   # Drive folder ID
       "Torts":     "2xYzAbCdEfGhIjKlMnOpQr",    # Drive folder ID
       "Civ Pro":   None,                          # no Drive sync
   }

On first run, a browser window will open to authorize access. After that, auth is saved automatically.

If you don't use Google Drive, set ENABLE_GOOGLE_DRIVE = False in config.py and skip this section entirely. Just drop files into the input folders manually.

---

Troubleshooting

ASSEMBLYAI_API_KEY environment variable is required Your .env file is missing or the key isn't set. Open .env and make sure the key is pasted on the correct line with no extra spaces.

ffmpeg not found or audio conversion errors FFmpeg isn't installed or isn't on your PATH. Re-run the FFmpeg install step and restart your terminal.

Class folder does not exist The class name in your CLASSES dict doesn't match a real folder inside PARENT_FOLDER. Check for typos — the name must match exactly, including capitalization and spaces.

Gemini API errors / model not available The model name in GEMINI_MODEL may be unavailable in your region or account tier. Try changing it to gemini-1.5-pro as a fallback.

No files processed (empty output) Make sure you dropped files into the correct input folder: open-law-notes/lecture-input/ for audio, open-law-notes/reading-input/ for readings. The folder names must match exactly.

---

License

MIT — see LICENSE.