# Memory

> Memory is a shared memory layer for AI tools and agents. It helps teams connect scattered work context from tools like Slack, Linear, GitHub, Gmail, documents, and AI conversations, then makes the right context available to ChatGPT, Claude, Claude Code, Cursor, and other AI agents.

Memory is built for teams and individuals who use AI tools heavily but repeatedly face the same problem: AI does not retain the right work context across tools, sessions, and agents.

Instead of forcing users to paste the same background information again and again, Memory.inc helps AI systems retrieve and use the relevant context they need.

## What Memory does

Memory connects work context from the tools people already use and turns it into a memory layer that AI tools can access.

This includes context from:

- team conversations
- documents
- emails
- GitHub issues and pull requests
- Linear tasks
- customer conversations
- AI chats
- product decisions
- internal rules and operating principles

The goal is to help AI tools understand the user, the team, the project, and the work history without requiring all context to be manually pasted into every prompt.

## Core idea Most AI tools are still session-based.

They can answer well inside one conversation, but they often lose context across sessions, tools, people, and agents.

Memory is designed to solve this by giving AI tools a persistent memory layer for work context.

This memory is not just raw storage. It needs to distinguish between:

- personal context and shared team context
- recent information and outdated information
- important information and noise
- private information and reusable team knowledge
- relevant context and unrelated data

## Context Retrieval

Memory is not designed to dump every piece of stored information into an AI prompt.

Instead, it retrieves the context that is most relevant to the current task. Relevant context can be found using signals such as:

- meaning
- keywords
- relationships
- source
- time
- usage patterns
- current task intent
- personal or team scope

This makes Memory different from simple document search or basic vector similarity search.

## Memory Decay

Not all information should stay equally important forever.

Memory is designed around the idea that older, less-used, or less-relevant information can become less prominent over time.

At the same time, old information can still remain important if it continues to be useful or repeatedly referenced.

This helps AI tools avoid overusing outdated, noisy, or irrelevant context.

## Who Memory is for

Memory is for people and teams that use AI tools as part of their real work.

Primary users include:
- AI-heavy software teams
- founders and operators
- product teams
- engineering teams
- consulting teams
- sales and customer-facing teams
- teams using ChatGPT, Claude, Claude Code, Cursor, Slack, Linear, GitHub, Gmail, and internal documents
- individuals who use multiple AI tools and want consistent memory across them

Memory is especially useful when important work context is scattered across many tools and needs to be reused by AI agents accurately.

## Key use cases

- Give Claude Code the right product and engineering context
- Help ChatGPT understand team decisions and customer history
- Let Cursor and other coding agents access relevant project memory
- Reduce repeated context-pasting across AI tools
- Connect scattered company knowledge into an AI-usable memory layer - Share memory across multiple AI agents
- Keep personal memory separate from team memory
- Retrieve only the right context instead of flooding the prompt with everything - Help AI tools maintain continuity across sessions

## What Memory is not

Memory is not just:
- a document search tool
- a company wiki
- a CRM - a notes app
- a generic RAG tool
- a vector database
- a chatbot wrapper

Memory is a memory infrastructure layer for AI tools and agents.

Its purpose is to help AI systems retrieve, reuse, and maintain the right work context across tools, sessions, people, and agents.

## Product links

- [Website](https://memory.inc/en/): Main English website for Memory.inc.
- [Machine view](https://memory.inc/en/machine): AI-readable overview of Memory.inc.
- [About](https://memory.inc/en/about): Company and product background.
- [Waitlist](https://memory.inc/en/waitlist): Early access registration.
- [Integration request](https://memory.inc/en/integration-request): Request support for additional tool integrations.
- [Contact](https://memory.inc/en/contact): Business, partnership, investor, or product inquiries.
- [News](https://memory.inc/en/news): Product updates, company news, and insights.