Overview of Rememberall: Long-Term Memory Solution for LLMs

Rememberall is an open-source project designed to enhance long-term memory storage capabilities for large language models (LLMs). It provides a secure and efficient way for AI developers and GPT Store builders to integrate persistent memory across conversations in custom GPT configurations. This tool is particularly useful for applications requiring the retention and retrieval of information over time, improving the continuity and contextuality of interactions.

Key Features

Privacy-First: Users have the option to self-host their memory store, ensuring control over data privacy and security.
Vector-Based: The system utilizes efficient semantic search mechanisms to manage and retrieve stored memories.
Easy Integration: Rememberall offers a simple REST API for integration, making it accessible for developers with varying levels of expertise.
Open Source: The project is open-source, allowing developers to customize and extend the system as needed.

How It Works

Rememberall operates through a straightforward three-step process:

Store Memories: The GPT can store essential information from conversations, such as details discussed or decisions made.
Load Past Context: When needed, the system can easily retrieve relevant memories to provide context to ongoing interactions.
Natural Interactions: The integration works seamlessly within the conversation flow, making interactions appear more natural and informed.

Example Usage

In a typical use case, when a user asks about a past conversation, the system uses Rememberall to recall the context:

User: "What did we discuss about authentication last week?"
Assistant: "Let me check @rememberall. According to our previous discussion, we implemented JWT-based auth..."

Quick Start Guide

To get started with Rememberall, developers can deploy the system using Docker Compose:

git clone https://github.com/yourusername/rememberall.git
cd rememberall/deploy
docker-compose up -d

Available API Endpoints

List/Search Memories: GET /memories?search=query&limit=10&offset=0
Create New Memory: POST /memory { "memory": "Your memory text here" }

Authentication

All API endpoints require Bearer token authentication. Tokens must be included in requests as follows:

Authorization: Bearer your-jwt-token

API Response Examples

Get Memories:

{
  "success": true,
  "memories": [
    {
      "id": "mem_123",
      "memory": "Discussion about authentication systems"
    }
  ]
}

Create Memory:

{
  "success": true,
  "memory": {
    "id": "mem_124",
    "memory": "New project requirements discussion"
  }
}

Architecture & Security

Rememberall is built with a focus on security and efficient data handling:

End-to-end Encryption: Ensures that all data stored and retrieved is fully encrypted.
Self-hosted Vector Storage: Allows for private, scalable storage solutions.
Fine-grained Access Control: Provides detailed control over who can access different types of data.
GDPR-compliant: Adheres to GDPR standards for data handling and privacy.

Rememberall is a practical solution for developers looking to enhance the capabilities of their LLMs with long-term memory functionalities, ensuring more dynamic and context-aware applications.