watch the live demo here

this project using the power of LLMs with the approach of Agents to provide an accurate and effective answer to users' queries.

• a Chat interface for interacting with the agent
• the agent has the ability to search the jewish library, read a specific text or get commentaries for a specific verse
• Flexible LLM provider support (Claude, GPT, Ollama)

• Python 3.11 or higher
• At least one of the following API keys (stored in .env):
  • Anthropic API key (for Claude)
  • OpenAI API key (for GPT)
  • Google API key (for Gemini) Or a Local Ollama setup (for open-source models).

1. Clone the repository
2. Install dependencies:

pip install -r requirements.txt

3. Set up a .env file with your credentials:

ANTHROPIC_API_KEY=your_anthropic_api_key
OPENAI_API_KEY=your_openai_api_key
GOOGLE_API_KEY=your_google_api_key

configure the index path:

INDEX_PATH=path/to/your/index

Run the Streamlit UI to see the system in action:

streamlit run app.py

The system uses a reason and act (ReAct) architecture, to achive an aoutonomous agent.

The agent can use the search engine to find relevant passages in the Jewish Library, and use those passages to answer user queries.

Key dependencies include:

• langchain and related packages for LLM integration
• streamlit for the user interface
• tantivy for document indexing and search
• python-dotenv for environment management
• Various LLM provider packages (anthropic, openai, ollama)

• Modify tantivy_search.py to customize search settings
• Modify sefaria.py to configure Sefaria API access
• Adjust agent.py to configure agent behavior
• Update llm_providers.py to add or modify LLM providers
• Customize app.py for UI modifications

• Store API keys and sensitive data in environment variables
• Never upload your .env file to version control
• Ensure proper access controls for the Tantivy index

• Verify the Tantivy index exists and is accessible
• Check API key permissions for your chosen LLM provider
• Ensure proper file permissions for the document library
• Review logs for detailed error information

Contributions are welcome! Please feel free to submit pull requests or open issues for bugs and feature requests.

MIT