Chat With Memory

A Python-based CLI chat application that demonstrates advanced conversation capabilities with persistent memory. Built using the Atomic Agents framework, this application showcases how to create intelligent agents that can remember previous interactions and form meaningful memories about conversations.

NOTE: This codebase is part of a tutorial article on integrating memory into your agents check it out here

Features

• 🧠 Memory Formation: Automatically forms and stores relevant memories from conversations using a dedicated Memory Formation Agent
• 🔍 Context-Aware Responses: Uses ChromaDB for semantic search and retrieval of previous memories
• 📅 Time-Aware: Includes current date context in conversations through context providers
• 💬 Rich CLI Interface: Beautiful command-line interface using Rich
• 🤖 Multi-Agent Architecture: Leverages Atomic Agents framework for sophisticated agent interactions

Prerequisites

• Python 3.10 or higher
• Poetry for dependency management

Installation

1. Clone the repository
2. Install dependencies using Poetry:

poetry install

3. Set up your environment variables in .env:

OPENAI_API_KEY=your_api_key_here

Usage

1. Activate the Poetry environment:

poetry shell

2. Run the chat application:

python -m chat_with_memory.main

3. Start chatting! The assistant will:
   • Respond to your messages using the Chat Agent
   • Form memories about important aspects using the Memory Formation Agent
   • Use previous memories to provide context-aware responses
   • Make intelligent choices about responses using the Choice Agent

To exit, press Ctrl+C.

Project Structure

chat_with_memory/
├── agents/                 # AI agents implementation
│   ├── chat_agent.py      # Handles main conversation flow
│   ├── memory_formation_agent.py  # Forms and manages memories
│   └── choice_agent.py    # Makes decisions about responses
├── tools/                 # Memory management tools
│   ├── memory_store_tool.py    # Stores memories in ChromaDB
│   ├── memory_query_tool.py    # Queries stored memories
│   └── memory_models.py        # Pydantic models for memory
├── services/              # Core services
│   └── chroma_db.py      # ChromaDB vector store implementation
├── main.py               # Application entry point
└── context_providers.py   # Provides time and memory context

Technical Details

Agents

• Chat Agent: Manages the main conversation flow and user interactions
• Memory Formation Agent: Analyzes conversations and forms relevant memories
• Choice Agent: Makes decisions about appropriate responses based on context

Tools

• Memory Store Tool: Handles the storage of memories in ChromaDB
• Memory Query Tool: Performs semantic search on stored memories
• Memory Models: Defines Pydantic models for memory structure

Core Technologies

• Atomic Agents: Framework for building and managing intelligent agents
• OpenAI API: Powers the language understanding and generation
• ChromaDB: Vector database for efficient memory storage and retrieval
• Pydantic: Data validation and settings management
• Rich: Terminal UI rendering
• Instructor: Enhanced OpenAI function calling

Dependencies

Key dependencies include:

• atomic-agents ^1.0.15
• rich ^13.9.4
• instructor ^1.6.4
• openai ^1.54.4
• pydantic ^2.9.2
• chromadb ^0.5.18

License

MIT License