Intro To Agents 12 2 2025

Meeting Notes Agent

An AI-powered agent that transcribes meeting audio, extracts action items, and maintains organized meeting notes in Markdown format.

🚀 **New here?** Start with [START_HERE.md](START_HERE.md) for a 60-second overview, or jump to [QUICKSTART.md](QUICKSTART.md) for the 5-minute setup guide!

Features

• 🎙️ Audio Transcription: Automatically transcribe meeting recordings using OpenAI Whisper
• ✅ To-Do Tracking: Extract and track action items with GitHub-style checkboxes
• 📝 Markdown Notes: Generate well-structured meeting notes with summaries and full transcripts
• 📋 Cross-Meeting Tracking: View all pending to-dos across all meetings
• ✓ Status Updates: Mark to-do items as complete

Setup

1. Activate the virtual environment:

source env/bin/activate

1. Set your OpenAI API key:

export OPENAI_API_KEY='your-api-key-here'

1. Run the agent:

python meeting_notes_agent.py

Usage

Interactive Mode

Run the agent and interact via the command line:

python meeting_notes_agent.py

Available commands:

• process - Transcribe and process a meeting audio file
• todos - List all pending to-dos across all meetings
• Type any message to chat with the agent
• quit - Exit

Programmatic Usage

from meeting_notes_agent import MeetingNotesAgent

agent = MeetingNotesAgent()

# Process a meeting audio file
response = agent.process_meeting_audio(
    audio_file_path="path/to/meeting.mp3",
    meeting_name="Team Sync",
    date="2024-12-03"
)

# List all pending to-dos
response = agent.chat("Show me all pending to-dos")

# Mark a to-do as complete
response = agent.chat("Mark the task 'Update documentation' as complete in the team sync meeting")

Meeting Notes Format

Meeting notes are saved in the `meeting_notes/` directory with the format: `YYYY-MM-DD_meeting_name.md`

Each note includes:

• Meeting name and date
• List of attendees
• Summar