Textbase is a Python framework for building AI chatbots, offering a simple interface for developers to integrate custom NLP/ML models, vector databases, and APIs. It targets developers seeking a flexible backend for chatbot creation, simplifying deployment and interaction logic.

How It Works

Textbase operates by abstracting the core chatbot interaction loop. Developers implement a single on_message function in main.py, which receives user input and returns a response. The framework handles message routing, state management, and potentially API calls, allowing developers to focus on the conversational logic and model integration. This approach provides flexibility by not dictating specific AI models or databases.

Quick Start & Requirements

• Install via pip: pip install textbase-client
• Local install: Clone repo, poetry shell, poetry install
• Requires Python >= 3.9.0.
• OpenAI API key is required for the default template.
• Start development server: textbase-client test or poetry run python textbase/textbase_cli.py test.
• Official docs: https://github.com/cofactoryai/textbase

Highlighted Details

• Simple on_message function for chatbot logic.
• Supports custom models, vector databases, and APIs.
• Easy web deployment via textbase-client deploy (coming soon).
• Local installation via Poetry for dependency management.

Maintenance & Community

Contributions are welcome via issues and pull requests. A contribution guide is available.

Licensing & Compatibility

The repository does not explicitly state a license in the provided README.

Limitations & Caveats

SMS integration and native support for models like Claude and Llama are listed as future features. The default template requires an OpenAI API key, implying potential vendor lock-in or cost implications for initial setup.