Email Tone Teller (ETT)

Email Tone Teller (ETT) is a smart assistant that reads client emails and helps you understand:

• The tone (friendly, annoyed, passive-aggressive, etc.)
• The urgency level
• Suggested replies
• Key action items

Designed for agencies, freelancers, support teams, and professionals who deal with high-volume, high-stakes email.

Planned Features

• Web interface to paste or upload email threads
• Slack and Teams integrations for tone analysis on demand
• Smart reply generator powered by LLMs
• Email-to-email relay system for live tone suggestions
• Admin dashboard for tracking sentiment over time (future)

Powered By

• Python (FastAPI backend)
• OpenAI / Claude / LLama via Openrouter (LLM Models)
• Mailgun (for live email)
• Slack Bolt SDK & Microsoft Graph (for bot integrations)
• React / HTMX frontend (optional phase)

Planned Structure

email-tone-teller/
├── backend/
│   ├── api-gateway/
│   ├── customer-api/
│   └── sentiment-api/
│       └── app/
│           ├── config/
│           │   └── llm_clients/
│           ├── dtos/
│           ├── routes/
│           ├── services/
│           ├── utilities/
│           └── main.py
│       ├── logs/
│       └── tests/
├── bots/
│   ├── slack/
│   ├── teams/
│   └── whatsapp/
├── frontend/
│   └── customer-web/
├── .env
├── .env.example
├── .gitignore
├── README.md
└── requirements.txt

Running the projects

python -m venv .venv
source .venv/bin/activate   # or .venv\Scripts\activate on Windows
pip install -r requirements.txt

Future Ideas

• Fallback logic (OpenAI → Claude → LLaMA)
• Slack + Teams notification integration
• Attachments summary (PDF, DOCX)
• Web frontend with email history