AutoOps

AutoOps is an automated operations system designed to streamline and automate email support responses based on a knowledge base built from previous interactions. The system uses AI to generate contextually relevant responses while allowing human review before sending.

Overview

AutoOps helps support teams by automating the initial response to customer emails. It processes incoming emails, searches a knowledge base for relevant information, and generates draft responses for review. The system learns from previous interactions to improve response quality over time.

Key Features

• Email Processing: Automatically fetches and processes incoming support emails
• Knowledge Base: Builds and maintains a searchable knowledge base from previous interactions
• AI-Powered Responses: Generates contextually relevant draft responses using OpenAI
• Vector Search: Uses embeddings for semantic search of the knowledge base
• Multi-language Support: Detects and responds in the customer's language
• Admin Dashboard: Web interface for reviewing and sending draft responses
• Real-time Processing: Webhook-based email processing for immediate handling

Tech Stack

Backend

• Node.js with Express
• TypeScript
• PostgreSQL with pgvector for vector storage
• Prisma ORM
• Gmail API for email integration
• OpenAI API for response generation and embeddings
• JWT for authentication

Frontend

• React with Next.js
• TypeScript
• Tailwind CSS

Architecture

Email Processing Pipeline

1. Email Fetching: Retrieves emails via Gmail API
2. Content Extraction: Parses email content and metadata
3. Language Detection: Identifies the language of the email
4. Knowledge Base Search: Finds relevant information using vector similarity
5. Response Generation: Creates draft responses using OpenAI
6. Human Review: Admin reviews and optionally edits responses
7. Sending: Approved responses are sent back to the customer

System Components

• Email Controller: Manages email fetching and processing
• Knowledge Controller: Handles knowledge base operations
• Response Controller: Manages response generation and sending
• OpenAI Service: Interfaces with OpenAI API for response generation
• Authentication Middleware: Handles user authentication and authorization

Current Status

• Backend core structure implemented
• Database schema defined with Prisma
• Email processing pipeline functional
• Response generation with OpenAI integrated
• Admin user management implemented
• Basic frontend dashboard operational
• Automated testing framework established
• Continuous operation system implemented

Getting Started

Prerequisites

• Node.js (v16+)
• PostgreSQL with pgvector extension
• Gmail API credentials
• OpenAI API key

Installation

1. Clone the repository

git clone https://github.com/GizAI/autoops.git
cd autoops

2. Install dependencies

# Install backend dependencies
npm install

# Install frontend dependencies
cd frontend
npm install
cd ..

3. Set up environment variables

cp .env.example .env
# Edit .env with your credentials

4. Set up the database

# Install pgvector extension in PostgreSQL
# Then run Prisma migrations
npx prisma migrate dev

5. Start the development servers

# Start backend server (from root directory)
npm run dev

# Start frontend server (in another terminal)
cd frontend
npm run dev

Project Structure

autoops/
├── .augment/         # Project documentation and task tracking
├── frontend/         # React Next.js frontend
│   ├── public/        # Static assets
│   ├── src/
│   │   ├── app/        # Next.js app router pages
│   │   ├── components/ # React components
│   │   └── lib/        # Utility functions and API clients
│   └── package.json
├── prisma/           # Prisma schema and migrations
├── src/              # Backend source code
│   ├── controllers/   # Request handlers
│   ├── middleware/    # Express middleware
│   ├── routes/        # API routes
│   ├── scripts/       # Utility scripts
│   ├── services/      # Business logic
│   ├── utils/         # Utility functions
│   └── __tests__/     # Test files
├── tests/            # E2E tests
└── package.json

Development Workflow

1. Email Processing Improvements:

   • Enhance email parsing and metadata extraction
   • Improve handling of attachments and complex email formats

2. Knowledge Base Enhancements:

   • Optimize vector search for better relevance
   • Implement knowledge base management features

3. Response Quality Improvements:

   • Refine OpenAI prompts for better responses
   • Implement response templates and customization

License

This project is licensed under the MIT License - see the LICENSE file for details.

Contact

For questions or support, please contact support@giz.ai