Meeting Transcript Summarizer

Convert meeting transcripts into structured notes. Extracts action items, key decisions, discussion points, and open questions. Supports Zoom .vtt, Teams transcripts, and plain text. Batch process entire folders.

What it does

Reads a meeting transcript file (plain text, WebVTT from Zoom/Teams, or SRT), sends it to Claude AI, and generates structured meeting notes with action items, key decisions, discussion points, and open questions. Meeting sentiment is also detected (productive, tense, neutral).

Features

• Parses .txt, .vtt (Zoom/Teams), and .srt transcript files
• Extracts action items with owner names
• Identifies key decisions and open questions
• Detects meeting sentiment and participants
• Batch process folders of transcripts
• Export to Markdown, printable HTML, or JSON

Quick start

git clone https://github.com/Get-Ai-Tools/meeting-summarizer
cd meeting-summarizer
pip install -r requirements.txt

# Summarize a plain text transcript
python summarizer.py --api-key sk-ant-... --transcript meeting.txt

# Summarize a Zoom .vtt file
python summarizer.py --api-key sk-ant-... --transcript meeting.vtt

# Batch process a folder, export HTML
python summarizer.py --api-key sk-ant-... --folder ./meetings --output html

Getting transcripts from video calls

• Zoom: Enable Cloud Recording → download the .vtt transcript file
• Google Meet: Use Tactiq or similar Chrome extension
• Microsoft Teams: Recordings include auto-generated transcripts

Requirements

• Python 3.8+
• anthropic (auto-installed)
• Claude API key - get one free at console.anthropic.com