AI-Powered Customer Support Chatbot 🤖

An intelligent AI assistant for customer support, trained on product manuals, FAQs, and ticket history, providing contextual answers using RAG (Retrieval-Augmented Generation) with LangChain + ChromaDB and deployed on AWS ECS.

---

🚀 Features

• Contextual Answers: Accurate responses using RAG pipeline
• Multi-Source Knowledge: Trained on manuals, FAQs, and ticket history
• Interactive Frontend: React.js UI for seamless customer interaction
• Scalable Deployment: Backend on AWS ECS
• Extensible Architecture: Easily add more documents or knowledge sources

---

📂 Project Structure

backend/
├── agents/
│   └── agent_fetcher.py     # RAG + GPT-4 agent
├── api/
│   └── app.py               # Flask API backend
├── requirements.txt
├── Dockerfile
frontend/
├── src/
│   ├── App.js               # React UI
│   └── App.css
├── Dockerfile
.env.example                  # Environment variables
docker-compose.yml            # Local testing setup
README.md                     # Project documentation

---

🏗 Architecture

flowchart LR
    A[User - React.js UI] -->|Question| B[Flask API Backend]
    B -->|RAG Request| C[LangChain Pipeline]
    C --> D[ChromaDB Vector Store]
    C --> E[GPT-4 LLM]
    E -->|Answer| C
    C -->|Response| B
    B -->|Answer| A

Loading

• Frontend: React.js
• Backend: Flask API
• RAG Pipeline: LangChain + ChromaDB + GPT-4
• Deployment: AWS ECS / Docker

---

⚙️ Setup

1. Clone Repository

git clone https://github.com/hq969/customer-support-chatbot.git
cd customer-support-chatbot

2. Backend Dependencies

cd backend
pip install -r requirements.txt

3. Frontend Dependencies

cd ../frontend
npm install

4. Configure Environment

Copy .env.example → .env and add your OpenAI API key:

OPENAI_API_KEY=your-openai-api-key
CHROMA_COLLECTION_NAME=customer_support

5. Run Locally

# Backend
cd backend
python app.py

# Frontend
cd ../frontend
npm start

Open http://localhost:3000 to interact with the chatbot.

---

🚀 AWS ECS Deployment

1. Create ECS Cluster (Fargate recommended)
2. Push Docker Images to AWS ECR:

# Backend
docker build -t customer-support-backend ./backend
docker tag customer-support-backend:latest <aws_account_id>.dkr.ecr.<region>.amazonaws.com/customer-support-backend:latest
docker push <aws_account_id>.dkr.ecr.<region>.amazonaws.com/customer-support-backend:latest

# Frontend
docker build -t customer-support-frontend ./frontend
docker tag customer-support-frontend:latest <aws_account_id>.dkr.ecr.<region>.amazonaws.com/customer-support-frontend:latest
docker push <aws_account_id>.dkr.ecr.<region>.amazonaws.com/customer-support-frontend:latest

3. Create ECS Task Definitions

   • Backend Task: port 5000, environment variables from .env
   • Frontend Task: port 3000

4. Create ECS Services using the tasks

5. Configure Application Load Balancer if public access is needed

---

⚡ CI/CD (GitHub Actions)

• Automatically build and push Docker images on main branch
• Updates ECS services automatically

# .github/workflows/deploy.yml
name: ECS Deploy

on:
  push:
    branches:
      - main

jobs:
  deploy:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
      - uses: aws-actions/configure-aws-credentials@v2
        with:
          aws-access-key-id: ${{ secrets.AWS_ACCESS_KEY_ID }}
          aws-secret-access-key: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
          aws-region: us-east-1
      - run: |
          docker build -t customer-support-backend ./backend
          docker tag customer-support-backend:latest <aws_account_id>.dkr.ecr.us-east-1.amazonaws.com/customer-support-backend:latest
          aws ecr get-login-password --region us-east-1 | docker login --username AWS --password-stdin <aws_account_id>.dkr.ecr.us-east-1.amazonaws.com
          docker push <aws_account_id>.dkr.ecr.us-east-1.amazonaws.com/customer-support-backend:latest
      - run: |
          docker build -t customer-support-frontend ./frontend
          docker tag customer-support-frontend:latest <aws_account_id>.dkr.ecr.us-east-1.amazonaws.com/customer-support-frontend:latest
          docker push <aws_account_id>.dkr.ecr.us-east-1.amazonaws.com/customer-support-frontend:latest
      - run: |
          aws ecs update-service --cluster customer-support-chatbot-cluster --service backend-service --force-new-deployment
          aws ecs update-service --cluster customer-support-chatbot-cluster --service frontend-service --force-new-deployment

---

💡 Benefits

• Cuts support response time dramatically
• Provides contextual and accurate answers
• Scales effortlessly with AWS ECS
• Easily extendable with new knowledge sources

---

👨‍💻 Author

Built by Harsh Sonkar ⚡

---