laminai Documentation

Everything you need to set up, configure, and get the most out of laminai. Start with the Quick Start below, or jump to the section you need using the sidebar.

Quick Start

Get laminai running in under 2 minutes:

1. Start the main appRun python3 app.py in the project directory. The app will be available at http://localhost:8080
2. (Optional) Start the landing pageRun python3 landing_app.py to serve the landing page at http://localhost:8000
3. Open the appNavigate to http://localhost:8080 in your browser
4. Paste a YouTube URLEnter any YouTube video URL and click "Analyze" — the app will download, transcribe, and analyze it automatically

Installation

laminai requires Python 3.8+ and the following packages:

pip3 install flask groq yt-dlp pydub pdfplumber python-dotenv

You also need FFmpeg installed for audio conversion:

brew install ffmpeg

sudo apt install ffmpeg

Configuration

Create a .env file in the project root with your API keys:

GROQ_API_KEY=gsk_your_key_here

Get your free API key at console.groq.com. The free tier provides generous limits for personal use.

YouTube Analysis

laminai processes YouTube videos through a 3-stage pipeline:

Progress is streamed in real-time via Server-Sent Events — you see download %, transcription progress, and analysis status live.

PDF Analysis

Upload any PDF document for instant AI-powered analysis. Supported formats include research papers, textbooks, reports, and presentations.

1. Click "Upload Document" in the sidebar or on the home screen
2. Select a PDF file from your computer (max recommended: 50MB)
3. Text is extracted using pdfplumber, preserving paragraph structure
4. Click "Analyze" to generate summaries, or go straight to AI Chat

AI Chat

laminai has two chat modes:

After processing a video or document, switch to the Chat view to ask questions grounded in the content. The AI uses the first 8,000 characters of the transcript as context.

Access the AI Assistant from the sidebar — no video or document required. This is a general-purpose chat interface, maintaining conversation history for the session.

MCQ Generator

After analyzing a video or document, click "Generate Quiz" to create multiple-choice questions. Configure the number of questions (5–20) from the settings panel.

Questions are distributed by difficulty: 30% Easy, 50% Medium, 20% Hard. Each question includes all 4 options, the correct answer, and a detailed explanation.

History

laminai automatically saves your activity to browser localStorage. The History view shows:

• YouTube videos — with video ID, timestamp, and quick-replay button
• Documents — PDF filename and upload date
• AI Chat sessions — first message as title, session timestamp

Up to 30 entries are stored. Use the filter tabs (All / YouTube / Documents / AI Chats) to browse by type. Click "Clear All" to reset.

API Endpoints

laminai exposes a REST API on http://localhost:8080. All endpoints return JSON except /api/process which streams SSE.

GET /api/process

Process a YouTube video. Streams progress events via Server-Sent Events.

Parameter	Type	Description
url	string	YouTube video URL (query param)

GET /api/process?url=https://youtube.com/watch?v=dQw4w9WgXcQ

// SSE stream events:
data: {"status":"downloading","message":"Downloading audio... 45%","pct":45}
data: {"status":"transcribing","message":"Transcribing audio...","pct":95}
data: {"status":"done","message":"Done!","pct":100,"transcript":"..."}

POST /api/process-document

Upload and extract text from a PDF file.

Body	Type	Description
file	multipart/form-data	PDF file to process

{"status": "done", "filename": "paper.pdf", "total_pages": 24}

POST /api/analyze

Analyze the current transcript (must call /api/process or /api/process-document first).

{"summary": "...", "key_topics": [...], "main_takeaways": [...],
 "important_details": [...], "practical_applications": [...]}

POST /api/generate-mcq

Generate MCQs from the current analysis.

Body (JSON)	Type	Default
num_questions	integer	10

POST /api/chat

Chat about the currently loaded video/document content.

Body (JSON)	Type	Description
message	string	User's question

POST /api/ai-chat

General-purpose AI chat (no content context required).

Body (JSON)	Type	Description
message	string	User's message
history	array	Prior messages [{role, content}]

Common Issues

# Find and kill the process using port 8080
lsof -ti:8080 | xargs kill -9
python3 app.py

You're likely using python instead of python3. Always use:

python3 app.py

Ensure your .env file is in the project root (same directory as app.py) and contains a valid key:

GROQ_API_KEY=gsk_xxxxxxxxxxxxxxxx

FFmpeg must be installed and in your system PATH. Verify with ffmpeg -version. If missing, install via Homebrew (macOS) or apt (Linux).

Support

If you're stuck on something not covered here, reach out:

• Email: support@laminaidigital.com
• Contact form: contact.html