Teams Interpreter Bot

A bot for Microsoft Teams that provides real-time translation and text-to-speech capabilities.

Features

• Translate messages between languages (English, Spanish, Russian)
• Convert text to speech using Windows TTS voices
• Bot commands for easy interaction
• Simple API for external integrations

Setup Instructions

Prerequisites

• Python 3.9+ installed
• Microsoft Azure account for Bot Registration
• Microsoft Teams account with permissions to add custom apps

1. Install Dependencies

pip install -r requirements.txt

2. Register a Bot in Azure

1. Go to the Azure Portal
2. Create a new "Bot Channels Registration" resource
3. Fill in the required details:
   • Name: TeamsInterpreterBot
   • Messaging endpoint: The public URL where your bot will be hosted (we'll use ngrok for testing)
   • Create a Microsoft App ID and password
4. Save the App ID and password

3. Configure Environment Variables

1. Open the .env file in the project root
2. Add your Microsoft App ID and password:

   MICROSOFT_APP_ID=your-app-id
   MICROSOFT_APP_PASSWORD=your-app-password
   

4. Start the Translation Server

python translation_tts_server.py

This starts the translation and TTS service on port 8080.

5. Create a Public Endpoint (for development)

To expose your local bot to Microsoft Teams, you can use ngrok:

ngrok http 3978

This will create a public URL like https://abc123.ngrok.io that forwards to your local server.

6. Update the Messaging Endpoint

1. Go back to your Bot registration in Azure
2. Update the Messaging endpoint with your ngrok URL + /api/messages (e.g., https://abc123.ngrok.io/api/messages)

7. Update the Teams Manifest

1. Open static/manifest.json
2. Replace YOUR_BOT_ID_HERE with your Microsoft App ID
3. Update validDomains to include your ngrok domain

8. Start the Bot Server

python app.py

This starts the bot server on port 3978.

9. Package and Upload to Teams

1. Create a zip file containing:
   • manifest.json (from the static folder)
   • color.png and outline.png icons (you can use placeholder images for testing)
2. In Microsoft Teams, go to "Apps" > "Manage your apps" > "Upload a custom app"
3. Upload the zip file

Bot Commands

• /help - Show help information
• /languages - Show supported languages
• /translate <text> - Translate text to another language
• /speak <text> - Convert text to speech
• /call <meeting-link> - Join a meeting (Coming soon)

Architecture

The system consists of three main components:

1. Translation and TTS Server - A Python HTTP server that handles translation and text-to-speech requests.
2. Teams Bot - A Bot Framework application that integrates with Microsoft Teams for messaging.
3. Calling Handler - Manages real-time media sessions for meeting interpretation.

Calling Features

The bot now supports joining Teams meetings to provide real-time interpretation. This requires:

1. The calling webhook configured in your bot registration
2. Additional Graph API permissions for calling
3. Proper Teams manifest settings with supportsCalling and supportsVideo enabled

For full calling functionality, you'll need to:

1. Configure your Bot Channel Registration with call capabilities
2. Add calls.accessMedia.all and calls.joinGroupCalls.all permissions to your app registration
3. Make sure your web server has the necessary endpoints to handle call events

Development

Running the Bot Locally

1. Start the translation server:

   python translation_tts_server.py
   

2. In a separate terminal, start the bot server:

   python app.py
   

3. Use ngrok to create a public URL:

   ngrok http 3978
   

Troubleshooting

• If the bot doesn't respond, check that your App ID and password are correct in the .env file.
• If translations are not working, ensure the translation server is running on port 8080.
• For TTS issues, ensure your Windows system has speech voices installed.

License

MIT License

Credits

Created by Your Name / Your Company.