https://github.com/alokthedataguy/multi-agent-customer-service-chatbot
This project is a multi-agent customer service chatbot designed for an e-commerce platform. The chatbot employ specialized agents handle distinct tasks to ensure efficient and accurate interactions. The chatbot aims to enhance user experience by streamlining order processing, answering FAQs, and providing personalized recommendations.
https://github.com/alokthedataguy/multi-agent-customer-service-chatbot
agentic-ai-development agentic-workflows chatbot customer-service docker e-commerce gradio-python-llm llm multi-agent-systems ollama
Last synced: 6 months ago
JSON representation
This project is a multi-agent customer service chatbot designed for an e-commerce platform. The chatbot employ specialized agents handle distinct tasks to ensure efficient and accurate interactions. The chatbot aims to enhance user experience by streamlining order processing, answering FAQs, and providing personalized recommendations.
- Host: GitHub
- URL: https://github.com/alokthedataguy/multi-agent-customer-service-chatbot
- Owner: AlokTheDataGuy
- Created: 2025-03-19T07:52:55.000Z (7 months ago)
- Default Branch: main
- Last Pushed: 2025-03-28T01:34:53.000Z (6 months ago)
- Last Synced: 2025-03-28T02:32:10.964Z (6 months ago)
- Topics: agentic-ai-development, agentic-workflows, chatbot, customer-service, docker, e-commerce, gradio-python-llm, llm, multi-agent-systems, ollama
- Language: Python
- Homepage:
- Size: 1000 Bytes
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: docs/README.md
Awesome Lists containing this project
README
# Customer Service Chatbot (WIP)
## Overview
This is a **modular, agent-based customer service chatbot** for an e-commerce platform. The chatbot is designed to handle various customer interactions, including order processing, FAQs, and personalized recommendations. It is built using **FastAPI, LangChain, Llama3:1B, Ollama, PostgreSQL, FAISS, Docker, and Gradio**.
## Features
- **Guard Agent**: Filters and blocks inappropriate or irrelevant queries.
- **Order Taking Agent**: Uses chain-of-thought reasoning to guide customers through the order process.
- **Details Agent (RAG System)**: Retrieves and generates responses about menu details, allergens, and FAQs.
- **Recommendation Agent**: Provides personalized product recommendations based on user orders.
- **Classification Agent**: Routes queries to the appropriate agent based on intent classification.
## Tech Stack
- **Backend:** FastAPI
- **NLP & LLM:** LangChain, Llama3:1B, Ollama
- **Database:** PostgreSQL
- **Vector Store:** FAISS
- **Containerization:** Docker
- **Testing UI:** Gradio
## Architecture
The chatbot follows a **modular agent-based architecture**, where each agent has a specific role. The agents work together in a **pipeline architecture** to ensure smooth interactions.
## Installation
### Prerequisites
- Python 3.9+
- Docker
- PostgreSQL
### Steps
1. Clone the repository:
```sh
git clone https://github.com/your-repo/customer-service-chatbot.git
cd customer-service-chatbot
```
2. Install dependencies:
```sh
pip install -r requirements.txt
```
3. Set up PostgreSQL and FAISS.
4. Run the chatbot using Docker:
```sh
docker-compose up --build
```
5. Access the Gradio UI for testing at `http://localhost:7860`
## Work in Progress
- Implementing **real-time query classification** for better routing.
- Fine-tuning **Llama3:1B** for improved response accuracy.
- Enhancing the **Recommendation Agent** with more personalization.
- Adding support for **multi-turn conversations**.
## Contributing
Contributions are welcome! Please fork the repository and create a pull request.
## License
MIT License
---
Stay tuned for updates as we continue to improve the chatbot!