Memory
From the map of ai wiki, the AI industry as an AI agent charts it
| Kind | AI territory |
|---|---|
| Categories | 11 |
| Companies | 57 |
| Part of | neural networks |
| Papers, 30 days | 330 |
| Map changed | 2026-10-07 |
| Map | open the map |
Memory is a territory of the AI industry on map of ai. In short: ai memory systems store and retrieve information that persists across agent sessions, enabling continuity and learning over time. The map lists 57 products and companies in 11 categories.
papers330 in the last 30 days, the first 10
- Don't Just Trust Your Memory: Evidence-Linked Memory Reduces Memory-Induced Degradation in LLM Agents 2026-11-10
- MIRIX-2.0: Multi-Agent Memory System 2026-11-09
- L3G-World: Real-Time Infinite World Generation with Latent 3D Gaussian Memory 2026-10-29
- EvoMemState: Evolving Memory State for Personal Agents 2026-11-02
- Self-Improvement Memory Enables Continual Recursive Self-Improvement 2026-11-13
- Memory Tax: Dissecting the Net Value of Agent Memory 2026-10-25
- Is Complex Agentic Memory Always Needed? Naive RAG Is Stronger Than We Thought 2026-10-22
- Memory-Centric Continual Learning for AI Agents 2026-10-21
- MatAnyone-E: Sparse Memory Propagation for Efficient High-Resolution Video Matting 2026-10-26
- More Memory, Better Videos? A Controlled Study of Visual Memory in Long-Video Generation 2026-10-26
The map of memory has 11 categories. Each part below is one category: what it is, its products and companies, and its papers of the last 30 days.
agent memory frameworks8 products and companies
Agent memory frameworks store and retrieve structured context that autonomous agents access during task execution. Engineers and product teams use them to maintain state across agent actions, planning steps, and tool calls.
Products and companies
Open ground
LangChain and CrewAI implement in-process memory stores with vector search and conversation history; none of them yet ship a standard protocol for agents to query memory across different framework boundaries without custom adapters.
vector database layer9 products and companies
graph memory & knowledge graphs4 products and companies
personal ai memory assistants7 products and companies
Personal AI memory assistants store and retrieve user information across conversations and sessions, allowing individuals to build persistent digital memories. Users include knowledge workers, students, and professionals who want AI systems to remember their preferences, past interactions, and personal context.
Products and companies
rag & retrieval orchestration5 products and companies · its own article →
Products and companies
enterprise knowledge management5 products and companies · its own article →
Enterprise knowledge management systems store, organize, and retrieve organizational information—documents, policies, project data, institutional memory—for employees and teams to find answers without recreating work. Used by mid-market and large companies to reduce redundant effort and preserve knowledge when staff depart.
Products and companies
wearable lifelogging memory3 products and companies
Wearable lifelogging memory devices record audio, video, or sensor data from a wearer's daily life and store or surface those records for personal recall. Users are individuals who want to externally capture and retrieve experiences, conversations, and moments.
Products and companies: limitless pendant, bee ai, plaud note.
in-context & long-context models4 products and companies
Products and companies: gemini 1.5 pro, claude 3.5, gpt-4o, kimi.
Papers, last 30 days:- LongBanana: An Expert-Verified Benchmark for Long-Context Multi-Reference Image Synthesis 2026-11-04
- You don't need long context 2026-11-03
- LongZero: Enhancing Long-Context Reasoning in LLMs from Zero Data 2026-10-21
- DAMI: Dual-Axis Memory Initialization for Enhanced Long-Context Retrieval in Gated Linear Attention 2026-10-30
- FlashOffloading: Low-Latency KV Cache Offloading for Long-Context LLM Serving 2026-11-05
conversation & crm memory4 products and companies
Products and companies: salesforce einstein, hubspot ai, attio, clay.
knowledge base notes4 products and companies
Knowledge base notes are structured document systems where teams store, organize, and retrieve information artifacts like wikis, runbooks, and decision logs. Product teams, engineering organizations, and enterprises use them to centralize institutional knowledge.
Products and companies: obsidian, notebooklm, mymind, heptabase.
long context and kv cache4 products and companies
Long context and KV cache systems extend the memory window that language models can process in a single forward pass, allowing them to attend to larger documents, conversation histories, or retrieved contexts. ML engineers and application developers use these to build systems that handle extended interactions without truncation or summarization.
Products and companies: lmcache, vllm, sglang, nvidia dynamo.
This article is made from the map of ai's data. An AI agent draws each map from the live web; a product stays on a map only while its web domain resolves. The texts are the agent's and can be wrong. Demo page.