This project provides a user interface for interacting with the OpenAI Assistants API, targeting developers and researchers who need a quick way to test and utilize the Assistants API's capabilities. It simplifies the process of setting up and managing conversations with AI assistants, including file uploads and streaming responses.

How It Works

The UI is built using Streamlit, a Python framework for creating data apps. It leverages the OpenAI Python client library to communicate with the Assistants API. Users configure their Assistant IDs and API keys via environment variables or a .env file. The application supports both single and multiple assistant profiles, allowing users to switch between different AI agents. It also includes optional authentication via Streamlit's streamlit-authenticator library.

Quick Start & Requirements

• Install dependencies: poetry install
• Run the app: streamlit run app.py
• Alternatively, use Docker: docker compose build then docker compose up
• Requires OpenAI API key and Assistant ID(s).
• Supports Azure OpenAI Service with appropriate endpoint and key configuration.
• Deployment to Streamlit Cloud is supported.

Highlighted Details

• Supports file upload and download within conversations.
• Implements streaming API responses for real-time interaction.
• Allows configuration for multiple assistant profiles.
• Compatible with Azure OpenAI Service.

Maintenance & Community

The project is maintained by ryo-ma. Further community or maintenance details are not explicitly provided in the README.

Licensing & Compatibility

The repository does not explicitly state a license in the provided README. Compatibility for commercial use or closed-source linking is not specified.

Limitations & Caveats

The project notes that Azure OpenAI Service support for streaming API is a prerequisite for its functionality with Azure. The README does not detail specific version requirements for Python or other dependencies beyond what poetry manages.