Aven AI Customer Support Agent 🤖

An AI-powered customer support agent for Aven, designed to answer user questions via Natural Language using a vector Database. Built with Python, Pinecone, and TypeScript, this project demonstrates the integration of web applications with AI-driven Q&A systems.

✨ Features

• Vector Database: Scrapes online information about Aven and stores it in a Pinecone vector database for fast semantic search.
• Natural Language Q&A: Users can interact with the AI agent via a TypeScript web app supporting voice input.
• Accurate & Reliable Responses: Designed to provide relevant answers by leveraging AI embeddings and similarity search.
• Optional Enhancements:
  • Evaluation set for measuring accuracy, helpfulness, and citation quality.
  • Guardrails for handling sensitive queries (personal, legal, financial, or toxic content).
  • Tool integration for scheduling meetings (optional).

🛠 Tech Stack

• Backend / AI: Python, Gemini API , Pinecone API , Firecrawl API
• Frontend: TypeScript, React (or Next.js)
• Data Storage: Pinecone vector database
• Others: Vapi API for real-time chat

🚀 Setup & Installation

1. Clone the repository

git clone https://github.com/YoshaM09/AI-Customer-Support-Agent.git
cd AI-Customer-Support-Agent

2. Install dependencies

pip install -r requirements.txt

3. Configure environment variables

• .env file for API keys (OpenAI, Pinecone, etc.)

4. Run the application

• Backend: python main.py
• Frontend: npm run dev

🎬 Usage

• Open the web app in your browser.
• Interact with the AI agent via natural language to ask questions about Aven.

🤝 Contributing

• Contributions are welcome! Please submit a pull request or open an issue for suggestions.

📄 License

• This project is licensed under the MIT License.