✉️ AI Email Agent

A full-stack Email Assistant using Flask, Django, and Groq LLaMA 3.1.
The system can:

• Categorize emails

• Extract actionable tasks
• Generate draft replies with AI

• Allow user instructions for personalized responses
• Save AI-generated drafts

• Manage customizable prompt templates

• Browse inbox and run agent actions interactively

---

🚀 Features

✔ AI-Powered Email Processing

Uses Groq LLaMA 3.1 (70B or 8B) to:

• Categorize incoming emails
• Extract JSON-formatted tasks
• Draft intelligent email replies
• Combine user instructions with prompt templates

✔ Django Frontend

Provides a clean UI for:

• Viewing inbox emails
• Running the AI processing agent
• Editing prompt templates
• Saving and reviewing generated drafts

✔ Flask Backend API

Handles:

• Prompt storage (SQLite)
• Local inbox (mock data)
• AI model calls via Groq
• Draft storage
• Email processing pipeline (categorize + action extraction)

---

🧩 Project Structure

email-agent/
│
├── backend-flask/
│   ├── app.py              # Flask API server
│   ├── llm.py              # Groq LLaMA 3.1 integration
│   ├── db.py               # SQLite database helpers
│   ├── data/
│   │   └── mock_inbox.json # Fake inbox used for development
│   ├── seed_prompts.py     # Seed default prompt templates
│   └── .env                # Groq API key + config (not committed)
│
└── frontend-django/
    ├── manage.py
    ├── frontend_django/
    └── emails/
        ├── views.py        # Inbox, prompts, drafts, agent
        ├── urls.py
        └── templates/
            └── emails/
                ├── base.html
                ├── inbox.html
                ├── prompts.html
                ├── drafts.html
                └── agent.html

---

⚙️ Requirements

Backend (Flask)

Flask
Flask-CORS
python-dotenv
requests
groq

Frontend (Django)

Django>=5.0
requests

---

🔧 Setup Instructions

1️⃣ Clone the repository

git clone https://github.com/USERNAME/email-agent.git
cd email-agent

---

2️⃣ Setup Backend

Create & activate virtual environment

cd backend-flask
python -m venv venv
venv\Scripts\activate     # Windows
# OR
source venv/bin/activate # macOS/Linux

Install dependencies

pip install -r requirements.txt

Create .env

GROQ_API_KEY=your_groq_key_here
GROQ_MODEL=llama-3.1-70b-versatile
FLASK_PORT=5000

Initialize database

python seed_prompts.py

Run Flask backend

python app.py

---

3️⃣ Setup Django Frontend

Create & activate virtual environment

cd ../frontend-django
python -m venv venv
venv\Scripts\activate
# OR
source venv/bin/activate

Install dependencies

pip install -r requirements.txt

Apply migrations

python manage.py migrate

Run Django server

python manage.py runserver

---

🧪 Usage Workflow

1. Open the frontend

Visit:

http://127.0.0.1:8000/

2. Browse Inbox

Click on any email to view details and run analysis:

• Run Agent (categorize + extract)
• Open Agent Chat → chat with AI

3. Generate Replies

In Agent Chat:

• Choose prompt template
• Add instruction (optional)
• Run Agent
• Save Draft

4. Review Drafts

Navigate to the Drafts page to see all AI-generated emails.

---

🤖 AI Model Integration

The backend uses:

✔ Groq LLaMA 3.1 (Recommended)

• llama-3.1-70b-versatile
• llama-3.1-8b-instant (faster)

In llm.py, call_llm() safely handles:

• Model calls
• API errors
• Network failures
• Returning clean text to frontend

---

📌 Roadmap

These may be added later:

• Gmail / Outlook OAuth integration
• Real email fetching
• Automated email replies
• Semantic search over inbox
• Multi-agent workflows
• Streaming responses
• Dark mode UI theme

---

🛡 Security Notes

• Never commit .env or API keys
• Add these to your .gitignore:

*.env
venv/
__pycache__/
*.sqlite

---

🧩 License

MIT License