PACE

Personality Accentuating Conversational Engine (PACE) is a framework designed to enhance conversational AI systems by integrating personality traits, emotional intelligence, and long term memory (using mem0). It combines a Digital Companion with a Personal Assistant to give an integrated experience.

Installation

1. Clone the repository

2. Install dependencies:

   uv sync

3. Configure environment:

   cp .env.example .env

   Edit .env and add required API keys (see .env.example and src/pace/config/constants.py).

4. Run the application:

   uv run main.py

Configuration

Configuration is managed via the .env file and src/pace/config/constants.py. For detailed setup of memory backends and LLM providers, refer to:

• Mem0 Vector DBs: https://docs.mem0.ai/components/vectordbs/overview
• Mem0 Embedders: https://docs.mem0.ai/components/embedders/overview
• Mem0 Graph Store: https://docs.mem0.ai/components/graph-store/overview

LLM providers (Gemini, Groq, etc.) require API keys in .env.

Usage

Once configured, start PACE:

uv run main.py

The application will start with the configured personality and memory systems, ready for conversational interaction.

Project Structure

PACE/
├── src/pace/           # Main package
│   ├── config/         # Configuration management
│   ├── graph/          # Graph-based conversation logic
│   ├── llm/            # LLM wrapper and rate limiting
│   ├── memory/         # Memory management
│   └── utils/          # Utility functions
├── Assistant/          # Character definitions and chat logs
├── logs/               # Application logs
└── main.py            # Entry point

Troubleshooting

• Import errors: Ensure you're using uv run to execute the application
• API key issues: Verify all required API keys are set in your .env file
• Memory backend errors: Check that Milvus, Ollama, and Neo4j services are running
• Debug mode: Set PACE_DEBUG=1 in your .env file for verbose logging