Skip to main content

Getting Started with Memoer

Let's discover Memoer - a powerful memory management system for Large Language Models (LLMs).

What is Memoer?​

Memoer is a specialized memory management system designed for LLMs that provides:

  • Easy-to-use abstraction for memory management and optimization
  • Native compatibility with Vercel's AI SDK
  • State-of-the-art (SOTA) memory management techniques

Installation​

Get started by installing Memoer:

npm install memoer

Basic Usage​

Here's a simple example of how to use Memoer:

import { memoer, MemoryConfig, ConversationStrategy } from "memoer";

// Create a new memory instance
const memoryConfig = {
  id: "conversation-1",
  systemMessage: {
    role: "system",
    content: "You are a helpful assistant."
  },
  managers: {
    conversation: {
      // Optional: Configure conversation strategy
      strategy: ConversationStrategy.SLIDING_WINDOW,
      slidingWindowSize: 10 // Number of messages to keep in context
    }
  }
};

// Initialize the memory
const memory = memoer.createMemory(memoryConfig);

// Add messages to the conversation
memory.conversation.add({
  role: "user",
  content: "Hello, how are you today?"
});

// Get optimized conversation context (affected by strategy)
const context = await memory.conversation.getContext();

// Get full conversation history regardless of strategy
const fullHistory = await memory.conversation.getFullContext();

Key Features​

  • Easy Integration: Designed to be simple to implement in any LLM application
  • Vercel AI-SDK Support: Built directly on top of Vercel's AI SDK, the mainstream unified LLM API platform
  • SOTA Memory Management: Advanced memory techniques to optimize context management for LLMs
  • Flexible Strategies: Configure different memory strategies like sliding windows to suit your needs

Next Steps​

Explore the documentation to learn more about Memoer's capabilities and how to integrate it into your LLM applications.