/Bee-AI

AI-chatbot with access to various tools (API's and custom functions) and responds back to the user on whatsapp using Twilio

Primary LanguagePythonMIT LicenseMIT

Bee-AI Chatbot

image

Bee-AI is a chatbot built using Flask and OpenAI's GPT-3.5 language model, designed to provide assistance through text-based conversations. It utilizes various tools and APIs to offer a range of functionalities, including information extraction, web browsing, and mobile payment integration.

Features

  • Text-based Conversations: Users can interact with the chatbot via text messages.
  • Natural Language Processing: Bee-AI leverages OpenAI's GPT-3.5 model for natural language understanding and response generation.
  • Tool Integration: It incorporates various tools and APIs to perform tasks such as information extraction, web browsing, and mobile payment processing.
  • Continuous Conversation Loop: The chatbot maintains a continuous conversation loop, allowing users to engage in multiple exchanges.
  • Customizable System Message: Users receive a customizable system message outlining the chatbot's capabilities and instructions on usage.

Tools and APIs Used

  • Google Serper API: An API for accessing search engine results pages from Google.
  • Safaricom Daraja API: An API provided by Safaricom for mobile payment integration in Kenya.

Planned Additions

  • Enhanced Natural Language Understanding: Implement advanced techniques to improve the chatbot's comprehension of user queries.
  • Additional APIs: Integrate more APIs to expand the chatbot's capabilities, such as location-based services, weather forecasting, and language translation.
  • User Authentication: Implement user authentication mechanisms for personalized interactions and data security.
  • Multimedia Support: Enable support for multimedia content, such as images, videos, and audio files, in conversations.

Usage

To interact with Bee-AI, follow these steps:

  1. Clone the repository: git clone https://github.com/yourusername/Bee-AI.git
  2. Install dependencies: pip install -r requirements.txt
  3. Run the Flask server: python main.py
  4. Send messages to Bee-AI via HTTP POST requests to http://localhost:5000/chat.

Future Enhancements

In future updates, we plan to add the following features and integrations:

  • Integration with Twilio API for handling SMS messages.
  • Expansion of toolset to include additional functionality such as language translation and sentiment analysis.

Contributors

License

licensed under the MIT License