> For the complete documentation index, see [llms.txt](https://cleyrop.gitbook.io/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://cleyrop.gitbook.io/docs/v4.7-en/getting-started/principaux-concepts-cleyrop.md). # Main Cleyrop concepts The key concepts to know to understand how the platform works Cleyrop is a sovereign Data & AI platform that centralizes, transforms, analyzes, and leverages data within a secure, collaborative environment. *** ### Projects A project brings together all the elements related to a use case: datasets, Work Data, corpora, dataflows, notebooks, Agents, etc. Each project has an owner and members with specific permissions (read, write, administration). ### Datasets Datasets represent the structured data of the platform. They can come from an external datasource (API, database, S3, SFTP, etc.) or be created from a dataflow. Each dataset has a sensitivity, a classification (bronze, silver, gold), and a status (valid, error, etc.). ### Dataflows Dataflows are data transformation pipelines. They can be Spark-based (distributed processing) or Python-based (unit or unstructured processing). Each step in a dataflow can produce a dataset and can be versioned, tested, or automated. ### Catalog The catalog centralizes all datasets available on the platform. Users can explore metadata there, preview data (depending on their access level), and request permissions. ### The Druide, the Agents and The Factory The Druide is the platform's secure conversational assistant: it communicates in natural language, analyzes your documents, queries your datasets, and generates documents or charts. Agents are Druide configurations specialized for a business need: instructions, knowledge scope (datasets, document corpora, files) and Tools. They are created in the project's AI Studio or directly from The Druide. The Factory makes it possible to deploy custom AI applications and tools to enhance the platform's capabilities (via Docker images or MCP servers). {% hint style="warning" %} The creation of new AI Assistants is blocked: they will soon be decommissioned in favor of the **Agents**. Use the Tools **Cleyrop Corpus** and **Cleyrop Dataset** in your Agents to achieve the same uses (querying document corpora and datasets). {% endhint %} ### Governance and administration Platform Managers ensure governance: management of users, projects, clusters, repositories, datasources, and permissions. Cleyrop includes fine-grained access management (roles, sensitivities, approvals) and complete tracking of logs and notifications.