Chat Personas using LLMs and Neo4j

A CLI harness for interacting with OpenAI LLMs and Neo4j via LangChain.

This is a companion for the sample code in the free Neo4j && LLM Fundamentals course, using a simple INI file for configuration.

The harness builds on the code samples in this course to provide a simple configuration based implementation with customizable "Persona" classes via an INI file.

Executing

Modify the chat_config.ini to:

1. Set your OpenAI API Key.
2. Review the 3 sample chat personas in the "CustomChatPersonas" section. Modify these settings between runs to test responses with different LLM and prompt settings. Add your own as well! :)
3. Execute main.py OR follow instructions for running in Docker

Notes

1. The custom personas do not interact with Neo4j (currently)
2. The movie persona requires a pre-populated Neo4j movies database only if "vector_search_enabled" is true in chat_config.ini.
   • A sandbox database can be created when you register and take Neo4j && LLM Fundamentals course