> ## Documentation Index
> Fetch the complete documentation index at: https://docs.cognee.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Core Concepts in Cognee Cloud

> How Cognee Cloud organizes memory, what the knowledge graph is made of, and how the four memory operations fit together

Cognee Cloud turns what you give it into memory that agents and people can ask questions of. This page explains the ideas behind that: where memory lives, what it is made of, and the operations that write, read, improve and remove it. Each section links to the Cloud page that shows how to do it from the app or the [HTTP API](https://docs.cognee.ai/cognee-cloud/http-api.md).

## Where memory lives

| Concept | What it is |
| - | - |
| **Workspace** | Your hosted Cognee, with its own database, members, datasets and URL. The API calls it a **tenant**. See [Workspace Lifecycle](https://docs.cognee.ai/cognee-cloud/workspace-lifecycle.md). |
| **Dataset** | A named body of long-term memory. Uploads, integrations and agents all write into datasets, and access is granted per dataset. See [Datasets](https://docs.cognee.ai/cognee-cloud/ui/datasets.md). |
| **Knowledge graph** | What a dataset's content becomes: entities and the relations between them, plus the documents and chunks they came from. Recall searches it. See [Graph](https://docs.cognee.ai/cognee-cloud/ui/knowledge-graph.md). |
| **Session** | Short-term memory for one agent conversation: its questions, answers and tool calls. See [Sessions](https://docs.cognee.ai/cognee-cloud/ui/sessions.md). |

Each workspace keeps its data, its vector embeddings and its knowledge graph in its own managed PostgreSQL database. Cognee manages the language and embedding models, so there is no model key to bring. See [Architecture](https://docs.cognee.ai/cognee-cloud/cognee-cloud-architecture.md).

## What the graph is made of

When Cognee processes your data, it turns it into **DataPoints**: typed records such as documents, chunks, entities, entity types and summaries. Each one is stored as a graph node with an embedding, so it can be found both by meaning and by following its connections. Recall searches over all of them, not only the entities the [Graph](https://docs.cognee.ai/cognee-cloud/ui/knowledge-graph.md) page draws.

The entity types come from an **ontology**. By default Cognee uses generic types such as *Person*, *Organization* and *Concept*. With the Premium [auto-ontology](https://docs.cognee.ai/cognee-cloud/functionality/auto-ontology.md), Cognee builds entity types from your own data, and you can refine them in plain language. You can also upload your own ontology file and pass it when you add data. See [Configuration & Ontologies](https://docs.cognee.ai/cognee-cloud/functionality/configuration-and-ontologies.md#ontologies).

Cloud does not accept custom graph models. To steer what gets extracted, use an ontology or a custom extraction prompt.

## The four memory operations

| Operation | What it does | In Cloud |
| - | - | - |
| **Remember** | Adds content to a dataset and builds its knowledge graph in one step. | [Data Ingestion](https://docs.cognee.ai/cognee-cloud/functionality/data-ingestion.md#remember) |
| **Recall** | Answers a question from memory, choosing the best way to search on its own. | [Search & Recall](https://docs.cognee.ai/cognee-cloud/functionality/search-and-recall.md#recall) |
| **Improve** | Turns sessions into long-term memory and refreshes the dataset's summaries and conflict review. | [Session Distillation](https://docs.cognee.ai/cognee-cloud/functionality/session-distillation.md#improve) |
| **Forget** | Removes one item, a dataset's graph, or a whole dataset. | [Knowledge Processing](https://docs.cognee.ai/cognee-cloud/functionality/knowledge-processing.md#forget) |

The older operations still work as lower-level steps. **Add** uploads content without processing it, **cognify** builds the graph from what was added, and **search** runs one search type you pick. Use them when you want to upload many files first and build the graph once. `remember` is add and cognify in one call, and `recall` picks the search type for you.

## Organizing and scoping memory

**Datasets** are the main boundary. Each one has an owner, the person who created it, and can be shared with teams and people. Permissions decide who, and which agents, can read or write it, and an agent never reaches more than its owner can. See [Memory Permissions](https://docs.cognee.ai/cognee-cloud/functionality/memory-permissions.md).

**Node sets** are tags inside a dataset. Pass `node_set` when you remember or add content, and later pass those names as `node_name` to limit a recall or search to that part of the graph. See [Scoping by node type](https://docs.cognee.ai/cognee-cloud/functionality/search-and-recall.md#scoping-by-node-type).

## Sessions and self-improvement

A session holds one agent conversation. Session entries are kept until a conversation has had no activity for one year.

Sessions become long-term memory through **improve**. It weights existing memories by the feedback in the session, saves the questions and answers, distills reusable lessons, and adds them to the dataset's graph, so the next session can recall what this one learned. When you connect a coding agent through [Agent Integrations](https://docs.cognee.ai/cognee-cloud/agent-integrations/index.md), the plugin runs this for you. In the app it is called **Self-improvement**.

The same run updates **personalized ranking**. Cognee reads how you rated earlier answers, both ratings you send and ones it infers from what you say next, and nudges your own future results toward what helped you. Your ratings never change a teammate's results. See [Personalized ranking](https://docs.cognee.ai/cognee-cloud/functionality/session-distillation.md#personalized-ranking).

## Moving memory

A dataset's knowledge graph can be exported as a **COGX** archive (`.cogx.tar.gz`), a portable file that another Cognee instance can import. See [COGX archive](https://docs.cognee.ai/cognee-cloud/functionality/dataset-management.md#cogx-archive) and [Syncing a local instance](https://docs.cognee.ai/cognee-cloud/connections/syncing-local-instance.md).

## Going deeper

Cognee Cloud runs the open-source Cognee engine. The [Cognee Core Concepts](https://docs.cognee.ai/core-concepts/overview.md) pages explain that engine in more depth, with examples for the Python library. Their code and server settings are for self-hosted Cognee and don't apply to your Cloud workspace.

> [card] [DataPoints](https://docs.cognee.ai/core-concepts/building-blocks/datapoints.md)
>
> The typed records the graph is built from.

> [card] [Remember](https://docs.cognee.ai/core-concepts/main-operations/remember.md)
>
> How content becomes graph-backed memory.

> [card] [Recall](https://docs.cognee.ai/core-concepts/main-operations/recall.md)
>
> How recall picks a search strategy.

> [card] [COGX](https://docs.cognee.ai/core-concepts/further-concepts/cogx.md)
>
> The format behind dataset export and import.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.