> For the complete documentation index, see [llms.txt](https://docs.oceanenterprise.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.oceanenterprise.io/developers/architecture.md).

# High-Level Architecture

Ocean Enterprise has a multi-layer architecture, as presented in the following diagram. 

<figure><img src="/files/eGnjzzK85J75OwjbVan9" alt=""><figcaption></figcaption></figure>

**Data Storage Layer:**  handles saving and retrieving of the data managed by the OE stack. The following types of storage are used:

* *IPFS* for storing the asset description. When an asset is created in OE, its description, including all metadata attached to the asset, is saved in IPFS.
* *Blockchain* for storing the reference to the asset description. After the description of the asset is created in IPFS, the reference to the IPFS object is saved in a transaction on the blockchain.
* *Web storage* for storing the actual content of the assets. For assets of type dataset or algorithm, their content is saved on a storage platform accessible via the HTTP protocol. The storage platform can be either on the publisher's premises or in the cloud.&#x20;

**Business Logic Layer:** Ocean Enterprise enables a decentralized exchange of data, on one side, and value, on the other side, between a publisher and a consumer. The core element that enables this exchange is represented by the Smart Contracts deployed on-chain.  To this end, the OE Smart Contracts implement the flows for publishing and consuming assets. This layer includes factory contracts, templates for data NFTs and data tokens, fixed-rate exchange contracts for paid assets, and Dispenser contracts for free assets.&#x20;

**Services Layer:** Represented by the Ocean Node component, this layer acts as a bridge between the Business Logic Layer and SDK, orchestrating how requests are processed and how business logic is executed. It provides the following services:

* orchestration of the entire consumption flows (download and Compute-To-Data).
* validation for requests.
* data encryption/decryption
* data streaming of the purchased asset to the consumer
* indexing mechanisms for assets
* abstraction of the complexity of the business and data layer, exposing clean APIs&#x20;

**SDK:** The ocean.js library encapsulates the Smart Contract functions to create assets as well as the services provided by Ocean Node in JavaScript functions, which can be used to develop OE-enabled business applications.&#x20;

**User Interface:** This facilitates the interaction between an end-user and the system. OE provides two interfaces:

* *Ocean Enterprise Marketplace*: a graphical interface where users can publish, retrieve, and consume assets in a very user-friendly manner. The user interface controls how data is displayed and how it responds to user actions. It also performs input validations before passing data to deeper levels.
* *Command Line Interface (CLI)*: a set of high-level tools that enable the creation and consumption of OE assets from a command line interface. This is appropriate when OE assets need to be manipulated in back-end like applications, where a user interface is not required. &#x20;

**Access Control Layer:** Is a critical component that governs who can interact with specific resources in OE and under what conditions.  It acts as a gatekeeper, enforcing policies that determine user permissions based on identity or other descriptive attributes. This layer controls, for instance, who is allowed to publish assets, who is allowed to consume a specific asset, or what algorithm is allowed to be executed on top of a specific dataset.