CoolFace
Apppublic

saadtoorx/multi-agent-customer-support-automation

sourceHugging Faceupdated 1y agoView on Hugging Face
0likes
App README

📞 Multi-Agent Customer Support Automation

An intelligent multi-agent customer support system powered by CrewAI and Streamlit. This application simulates how AI agents can collaboratively handle customer inquiries with high quality, empathy, and accuracy.

<div align="center"> <img src="images/demo_screenshot.png" width="80%" alt="App Demo Screenshot"> </div>

🚀 Features

  • —🤖 Multi-Agent Collaboration: AI agents perform specialized tasks such as resolving support inquiries and reviewing responses.
  • —🔍 Web Scraping Integration: Automatically pulls info from trusted sources like CrewAI Docs.
  • —💬 Natural, Friendly Tone: Outputs are helpful, casual yet professional — ideal for real-world support environments.
  • —🧠 Memory-Enabled CrewAI: Agents can retain shared memory to improve task performance.
  • —🧩 Modular Architecture: Easily extend or modify the system to include more agents or tools.

🏗️ Architecture Overview

🧠 Agents

  • —Support Agent: Responds directly to customer queries with researched answers.
  • —QA Agent: Reviews the support agent’s reply to ensure tone, accuracy, and completeness.

📋 Tasks

  1. 1.Inquiry Resolution: Resolves user inquiry with research and clarity.
  2. 2.QA Review: Reviews and polishes the initial response for final output.

🛠️ Tools Used

  • —WebsiteSearchTool – for general web searches
  • —ScrapeWebsiteTool – scrapes content from specific URLs
  • —Preconfigured doc scraper for CrewAI docs

📁 Project Structure

multi-agent-customer-support/
├── app.py                 # 🚀 Streamlit app interface
├── agents.py              # 🤖 Agent definitions (not shared yet)
├── tasks.py               # 📋 Task definitions
├── tools.py               # 🛠️ Tool configurations
├── utils.py               # 🔑 API key handling
├── requirements.txt       # 📦 Dependencies
├── README.md              # 📖 Documentation
└── images/
    └── demo_screenshot.png  # 🖼️ App UI screenshot (optional)

🛠️ Installation & Setup

  1. 1.Clone the repository
bash
git clone https://github.com/saadtoorx/multi-agent-customer-support.git
cd multi-agent-customer-support
  1. 1.Install dependencies
bash
pip install -r requirements.txt
  1. 1.Add your OpenAI API key
  2. 2.Create a .env file:
env
OPENAI_API_KEY=your-api-key-here

▶️ Running the App

bash
streamlit run app.py
  • —📝 App simulates a support inquiry from a fictional customer.
  • —🧠 CrewAI kicks off task execution between agents.
  • —📋 Final, polished response is displayed in Streamlit UI.

📦 Tech Stack

🤖 Agent Summary

AgentResponsibility
Support AgentResearches and writes a helpful reply to customer inquiry.
QA AgentReviews the reply, improves tone, and ensures completeness.

🔮 Future Improvements

  • —🧠 Add more agents (e.g., Sentiment Analyst, Follow-up Manager)
  • —🌐 Use live customer chat input instead of static prompt
  • —🧾 Export final replies to email or ticketing system
  • —🧩 Add memory visualization or trace logs

📄 License

Licensed under the MIT License. See `LICENSE` for full terms.

✨ Built By

Created with ❤️ by Saad Toor (@saadtoorx) Powered by CrewAI, Streamlit, and OpenAI