Introduction

CESS (Cumulus Encrypted Storage System) is a blockchain-powered decentralized cloud storage infrastructure. As the first decentralized data platform with its own Layer 1 blockchain, CESS offers virtually unlimited storage capacity integrated with ethical AI technologies. Leveraging its native Content Decentralized Delivery Network (CD²N), it enables millisecond-level data transmission, making it a comprehensive Web3 solution for storing and accessing high-frequency, dynamic data. With CESS, users and content creators can share data on-chain while preserving data sovereignty and user privacy. The platform empowers developers to build and deploy decentralized applications with secure, transparent, and high-throughput data management capabilities. CESS envisions a secure, efficient, and scalable decentralized cloud network—one that not only provides data storage and sharing services but also serves as an innovative solution to bring order to an increasingly chaotic digital world.

Project Milestones

2021
Launched testnet v0.1.

2022
Released testnet versions v0.1.2 through v0.6.
Launched blockchain explorer Substats v0.1.
Released Decentralized Object Storage Service (DeOSS).

2023
Released testnet versions v0.6.1 through v0.7.5.
Completed Substrate Builder Program.
Enhanced EVM and WASM contract compatibility.
Launched decentralized file-sharing tool DeShare.

2024
Proposed IEEE P3233 decentralized storage standard protocol.
Completed blockchain explorer Substats v2.0.
Released CESS Whitepaper v1.0.
Released CESS Economic Whitepaper v0.1.

2025
Launched Mainnet v1.0.
Launched CD²N Mainnet v1.0.
Released CESS AI-LINK component.

Team

Founded in 2019, CESS brings together international talent from the UK, US, India, Hong Kong, UAE, and Argentina. The team comprises cryptographers, data storage experts, and computer science engineers dedicated to advancing blockchain-based decentralized storage technology. Combining youthful energy with technical expertise and a passion for positive change, team members work to push technological boundaries and create meaningful social impact. Their core mission is to achieve excellence in digital technology through continuous innovation, delivering secure and efficient decentralized data storage and sharing solutions for the Web3 era.

Nicholas Zaldastani

Nicholas Zaldastani serves as CESS’s Chairman, Co-founder, and Head of Marketing. With extensive experience in technology, venture capital, and scaling companies, he previously served as a director at Oracle from 1988 to 1994, overseeing international marketing and product management. His Harvard Business School education and expertise in business strategy and growth bring exceptional leadership to CESS’s decentralized data value infrastructure development.

Joseph Li

Joseph Li serves as CESS’s Co-founder and Chief Technology Officer (CTO), focusing on decentralized cloud storage and Web3 data security. His expertise in cybersecurity and blockchain architecture plays a crucial role in developing CESS’s scalable and secure data sharing solutions.

Jessie Dai

Jessie Dai serves as CESS’s Co-founder and Chief Operating Officer (COO). She is a trader, entrepreneur, and early investor in cryptocurrency. As Vice Chairman of the Hong Kong Web3 Standardization Association, she actively contributes to Web3 technology development and implementation. Her background in blockchain strategy and ecosystem growth plays a vital role in CESS’s operations, partnerships, and industry engagement.

How Does it Work?

CESS’s core technical architecture consists of two main module systems: the CESS Protocol Suite and the XESS AI Protocol Suite. These modules are connected through an Interface layer, which facilitates interaction between internal elements and external systems.

CESS Protocol Suite

This forms the foundation of the CESS network, responsible for data storage, management, and distribution. It consists of three core layers:

1. Blockchain Layer

This layer forms the foundation of the entire network and delivers blockchain solutions. It primarily integrates idle storage and computing resources to enable data storage, verify data rights, and provide application services. The layer contains essential components—Consensus Nodes, Validator Selection (RPS), consensus algorithms, encryption systems (PRE), and virtual machines—which together ensure the network’s decentralization, security, and programmability.

2. Distributed Storage Resource Layer

This layer uses virtualization technology to integrate and pool distributed storage resources into a unified resource pool. Its infrastructure includes Storage Capacity Nodes and Storage Scheduling Nodes, which handle actual data storage and management tasks. To ensure data security and availability, this layer incorporates mechanisms such as Data Ownership (MDRC), Storage Proof (PoTS/PoDR), and Data Availability. The layer also features TEE (Trusted Execution Environment) nodes for enhanced data privacy and secure processing.

3. Content Decentralized Delivery Network Layer - CD²N

This layer is central to CESS’s high-speed data distribution capabilities. Using content caching technology, it ensures rapid data retrieval and distribution. The layer involves Data Index Nodes (known as Retrievers) and Data Delivery Nodes (known as Cachers). Retrievers locate data while Cachers provide quick access to data copies. To optimize distribution efficiency, the CD²N layer includes Traffic Algorithm (FDT), Load Balance, and Data Sovereignty (LBSS) mechanisms, ensuring efficient data distribution and user control over their data.

CESS Network features a carefully designed data storage workflow that offers intelligent processing for images, videos, and documents. This streamlines online data processing while giving users control over data removal. Through blockchain tracking of all operations, CESS ensures complete transparency and traceability.

When a user initiates a data storage request, CESS platform begins a preprocessing step. First, the CESS client software uploads and preprocesses the user’s data file. During this phase, the system extracts and stores the file’s metadata (such as data owner identity, keywords) and data fingerprint (for confirming data ownership). This metadata and fingerprint are then submitted to the CESS chain for recording. The preprocessing also manages file replication and applies fault-tolerant erasure coding.

After preprocessing, data files are split into smaller segments (Slice Files). The system then applies erasure coding to these segments. Users can customize the encoding rate based on the importance of data segments, meaning that even if some segment copies are corrupted, the original data can be recovered through fault-tolerant algorithms, greatly enhancing data availability and disaster recovery capabilities. The processed data fragments are then distributed to randomly selected storage nodes in the CESS storage network.

When data fragments arrive at storage nodes, the nodes request data tags from TEE Workers (with consensus nodes assisting in tag calculation). As shown in the diagram, each storage node receives corresponding tags (Tag1 to Tag5). These data tags are stored locally alongside the received file fragments. The tags contain validator signatures, making them tamper-proof and crucial for subsequent data integrity verification. After successfully storing the data and saving tags, storage nodes report their storage status to the CESS chain, marking the data file as reliably stored.

To ensure ongoing data integrity and storage node reliability, the CESS network employs periodic challenge procedures called Proof of Data Reduplication and Recovery (PoDR²). At irregular intervals, consensus nodes issue random challenges. In response, storage nodes must generate Proof of Data Integrity using their stored data fragments and associated tags, and submit these proofs for verification by TEE Workers within a defined deadline.

Storage nodes also regularly submit Proof of Data Possession to the CESS blockchain. Failure to complete a challenge and submit proof on time results in the affected data files being unrecognized by the CESS chain, and the responsible storage node facing penalties. For greater efficiency, storage nodes can batch-submit calculated proofs to the blockchain.

The PoDR² mechanism integrates erasure coding and Proof of Data Possession (PDP) technology. Erasure coding enhances data availability through redundancy, while the PDP process effectively deters dishonest behavior by verifying that data is truly stored and readily accessible.

XESS AI Protocol Suite

This module suite focuses on leveraging cutting-edge AI technologies to enable secure and private collaborative model training across the entire CESS network.

1. CESS AI Agent Hub

It provides a unified entry point for users and applications to access, connect, and deploy AI Agents across industries. By leveraging CESS network’s data advantages, the AI Agent Hub simplifies AI integration complexity while providing a decentralized, scalable, and secure AI infrastructure.

2. CESS AI-LINK

This is the core component of the XESS AI Protocol Suite. It integrates federated learning mechanisms, allowing participants to train shared models without sharing their raw data. AI-LINK uses smart contracts to delegate computational tasks to nodes across the network, ensuring efficient resource utilization while maintaining data sovereignty. This component significantly enhances the network’s AI capabilities, supporting complex AI applications and facilitating industry-wide collaboration without compromising data privacy.

Interface

The Interface Layer serves as a bridge in CESS’s architecture. It manages interactions and communications between different modules of the CESS Protocol Suite and XESS AI Protocol Suite, while defining a set of rules and conventions that enable various components to work together seamlessly, delivering CESS’s full functionality. Additionally, the Interface Layer facilitates the creation, management, and interaction with external blockchain networks and Web3 DApps through CLI, RPC, API, and SDK interfaces. This enables CESS to integrate smoothly into the broader Web3 ecosystem.

Technical features

Random Rotational Selection (R²S)

CESS utilizes a consensus mechanism known as Random Rotational Selection (R²S), which is designed to efficiently facilitate block production and manage on-chain transactions. R²S offers an open framework that allows users interested in becoming node operators to join a candidate node pool. Within fixed time windows (e.g., every 3,600 blocks), the system dynamically selects 11 rotating nodes from this pool to be responsible for block production. Candidate nodes that are not selected for block production are assigned auxiliary tasks, such as data preprocessing. This allows them to demonstrate their operational capabilities and increases their chances of being promoted to rotating nodes in future rounds.

R²S incorporates a credit scoring system that continuously evaluates node behavior and performance. Nodes found to be underperforming, engaging in malicious activities, or failing to meet network requirements are penalized with reduced credit scores. Nodes whose scores fall below a predefined threshold are disqualified from the candidate pool. Similarly, rotating nodes that act maliciously or fail to fulfill their responsibilities are promptly removed and replaced by new nodes randomly selected from the candidate pool. This ensures the protocol’s continuity and fairness. In terms of node entry and exit, CESS maintains a relatively open access policy. Participants must meet the basic operational and resource contribution standards required by the network and must stake a predetermined amount of $CESS tokens as collateral to mitigate the risk of malicious behavior. Upon exiting the network, a performance assessment determines whether the staked tokens will be refunded. Well-performing nodes receive a full refund, while those that remain offline for extended periods or engage in misconduct may forfeit part or all of their stake. This entry and exit mechanism incentivizes honest participation and strengthens network security by deterring potential attacks, thereby enhancing the stability of the consensus process.

Node election lies at the heart of block production under R²S. To become a consensus candidate, a node must stake 3 million $CESS tokens. In each rotation cycle, 11 validators (the rotating nodes) are selected based on their comprehensive scores, which include credit score, stake score, and VRF (Verifiable Random Function) score. Once selected, consensus nodes are not only responsible for maintaining network integrity but also carry out critical tasks such as data preprocessing and verifying file content and idle storage space during random challenges. They may also be required to certify or replace idle space. CESS motivates reliable participation through a credit-based evaluation system that assesses each validator’s contributions. These contributions directly impact the node’s credit score.

The R²S consensus mechanism offers several key advantages. First, by introducing randomized rotational selection, it effectively prevents monopolization and centralization, ensuring no single large node can unduly influence the network. Second, the rotation of 11 nodes per cycle for block production and verification boosts consensus efficiency while maintaining decentralization. Finally, R²S supports fast and efficient on-chain transaction processing, especially for metadata, enabling direct data storage addressing on the blockchain and ensuring data authenticity through blockchain-based verification.

Multiple Data Storage Proof Algorithm

In decentralized storage networks, incentivizing users to contribute idle storage resources presents a core challenge: how to ensure data integrity in the presence of potentially malicious behavior. Common threats include storage space fraud (where nodes falsely report their capacity) and outsourcing attacks (where colluding nodes store duplicate data under the guise of independent storage, undermining redundancy and reliability). While existing cryptographic mechanisms—such as Proof of Storage, Proof of Replication, and Proof of Space-Time—help verify storage claims and ensure secure, redundant data retention, some of these methods face scalability and efficiency limitations, particularly in high-frequency data retrieval scenarios.

To overcome these challenges and improve the reliability of its storage services, CESS introduces two innovative data storage proof techniques: Proof of Idle Space (PoIS) and Proof of Data Reduplication and Recovery (PoDR²). PoIS verifies the availability and integrity of idle space (i.e., segments not storing user data) provided by storage nodes; PoDR² verifies the integrity and possession of active user data (i.e., service data segments) stored by the nodes.

PoIS (Proof of Idle Space) addresses the challenge of accurately measuring and verifying unused storage space that is not occupied by user data. Since it’s not feasible to directly access disk contents as in traditional systems, PoIS requires nodes to fill their idle space with randomly generated “idle files.” These files are securely maintained using proof-of-storage mechanisms to ensure ongoing possession by the storage node. To improve efficiency, PoIS adopts a three-layer (or multi-layer) hierarchical accumulator structure, optimizing both space usage and computational performance. When an element in a sub-accumulator is updated, only its parent and relevant sibling accumulators need recalculation, reducing overhead. To prevent fraudulent behaviors such as compression, on-demand generation, or cross-validation, CESS utilizes a “stone-laying game” built on a Stacked Bipartite Expander Graph to generate and manage idle files securely. PoIS is a dynamic mechanism—nodes can manage their storage space flexibly and must respond to validator challenges to prove the integrity of their claimed idle space.

The Proof of Data Reduplication and Recovery (PoDR²) focuses on verifying that storage nodes reliably hold user data (i.e., service data segments). PoDR² combines two technologies: Erasure Coding (EC) and Proof of Data Possession (PDP). It ensures data availability by slicing user files, applying erasure coding to generate redundant data blocks, and distributing these fragments across multiple storage nodes. At the same time, PoDR² implements the PDP mechanism to prevent fraudulent behavior by storage nodes. Nodes must periodically submit proofs of data possession to the blockchain, based on the stored data fragments and tags generated by a Trusted Execution Environment (TEE). This process verifies data integrity and ensures that user data is reliably maintained. The periodic challenge process of PoDR² is a core component of the overall storage system. It ensures that storage nodes continuously fulfill their data retention responsibilities.

Use Cases

With its secure data infrastructure, the CESS network supports a wide range of use cases.

1. Data Availability Service (DA Service): The CESS network provides reliable data access services by replicating data across multiple nodes. This ensures data redundancy and fault tolerance, maintaining availability even in the event of network interruptions or node failures. Additionally, the DA Service can act as a Layer 2 storage solution for major blockchain networks such as Bitcoin and Ethereum. It helps offload large datasets from these networks, reducing on-chain storage costs and increasing transaction speed while preserving decentralized and secure data storage. Its scalability and robustness make it suitable for a wide range of applications, including decentralized finance (DeFi), enterprise storage, and large-scale data management.

2. Distributed Network Disk: CESS offers a unique distributed network disk service for end-users, delivering significant advantages over traditional cloud storage providers. By storing data across multiple independent nodes instead of centralized servers, it enhances security, data ownership, and storage capacity. This decentralized approach eliminates reliance on centralized services and enables faster upload and download speeds. With the use of blockchain and advanced encryption technologies, CESS guarantees data privacy and security, avoiding the risks of data loss associated with centralized servers. Furthermore, storage nodes can dynamically join the network and contribute idle space, enabling unlimited scalability of the storage network.

3. Distributed AI Training: CESS significantly enhances distributed AI training by offering secure and scalable storage for training data. The network’s high bandwidth and low latency ensure efficient data transmission between nodes, shortening training times. With CESS, AI developers can collaboratively train models while preserving data privacy and security through federated learning and encryption technologies. This addresses the common issues of data silos and privacy leakage in traditional AI training environments.

4. Decentralized Digital Assets Marketplace: In digital asset markets, secure storage, decentralization, and trust in transaction data are essential. CESS plays a key role in this scenario by verifying digital assets such as NFTs through its multi-format data rights confirmation mechanism. After developers or asset owners upload files to CESS for verification, the data is distributed across storage nodes. CESS can automatically capture the structural, thematic, and semantic features of digital assets to build a vector space, enabling precise indexing and mapping. This enhances public discovery capabilities and enables secure private retrieval, thereby increasing trust and efficiency in the digital asset marketplace.

Ecosystem

The CESS ecosystem is actively expanding its collaborative network, forging strong partnerships with major traditional tech giants such as AWS, Intel, and Tencent, as well as leading blockchain projects like Polkadot and IoTeX. In addition, numerous other initiatives and organizations, such as the Web3 Foundation, IEEE, and GBA, have become important ecosystem partners of CESS, jointly promoting the adoption and advancement of CESS technology. CESS has also earned industry recognition, including approval of IEEE standards, significantly enhancing its credibility and broadening its application potential. These achievements provide a solid foundation for the healthy growth of the CESS ecosystem.

In 2025, CESS formed a strategic partnership with GAIB, an organization focused on building the economic layer for AI computing through tokenized, income-generating GPU assets and its AI-synthesized dollar, $AID. As a complementary force, CESS delivers a high-performance, encrypted, and privacy-focused storage infrastructure to support dynamic datasets. This collaboration seamlessly integrates computing and storage resources, combining GAIB’s computational power with CESS’s robust storage framework. The partnership aims to improve the efficiency and security of AI and DeFi protocols, while jointly driving the development of decentralization.

At the same time, CESS plays a key role as a core member of the Hong Kong Web3.0 Standardization Association (W3SA), contributing significantly to W3SA’s 2025 conferences and summits. CESS researcher Tony Dai delivered a keynote speech on the standardization of decentralized physical infrastructure and the future of distributed storage evaluation. The speech highlighted CESS’s role as a founding member and initiator of IEEE P3220.02—the world’s first international standard for decentralized storage protocols based on blockchain. This standard is vital for the DePIN and RWA infrastructure stack, as it defines frameworks for data availability, recovery, auditability, DePIN network performance evaluation, and reputation scoring in decentralized environments, as well as cross-border data compliance through mechanisms like LBSS. CESS’s involvement in W3SA and its leadership in advancing industry standards—particularly in building trust, compliance, and interoperability infrastructure necessary for onboarding Real World Assets (RWA) onto the blockchain—further strengthens its position as a key player in the Web3 ecosystem.

Tokenomics

The tokenomics of CESS is based on a total supply of 10 billion CESS tokens. Of this supply, 15% is allocated to initial contributors, 10% to early investors, 10% for community development, incentives, and promotion, 5% for business partnerships with cloud service providers, and 5% is reserved by the foundation for emergencies and long-term ecosystem development.

The largest allocation—a substantial 55%—is dedicated to incentivizing nodes that support the storage network. Specifically, 30% is allocated to storage nodes, 15% to consensus nodes, and 10% to the development of the caching layer. This distribution reflects CESS’s strong emphasis on building a powerful and reliable decentralized storage infrastructure.

CESS tokens are the native cryptocurrency of the CESS network and play multiple vital roles within the ecosystem. They serve as a medium for staking to earn passive income, grant holders the right to participate in governance, and are required to access various storage services across the network—functioning as the key to CESS’s decentralized storage capabilities.

Storage nodes earn rewards for contributing storage space, providing data hosting and download services, and performing data validation tasks. These rewards include mining incentives and a portion of storage service fees. The amount of tokens a storage node must stake is based on its declared storage capacity. Nodes must regularly complete random challenges—Proof of Idle Space (PoIS) for verifying unused space and Proof of Data Reduplication and Recovery (PoDR²) for verifying user data—to prove both the authenticity and reliability of their contributions. The rewards distributed to storage nodes are proportional to their “power” within the network, which reflects their share of total verified storage capacity. In each reward cycle, a fixed number of tokens is distributed based on this power ratio. Storage nodes may exit the network at any time, but they are required to assist in data migration to ensure the safety of user data. If a node repeatedly fails to complete random challenges—due to downtime, disconnection, or data loss—it will be forcibly removed from the network, and its staked tokens will be partially or fully slashed as a penalty.

Risk Analysis

Although CESS is designed with a strong emphasis on security and efficiency at both the technical and economic levels, it still faces several inherent risks as a decentralized network.

First, storage nodes may be motivated to engage in malicious behavior, such as falsifying their claimed Proof of Idle Space (PoIS). To counter such threats, CESS employs a combination of technical safeguards—including PoIS, random challenges, and verification mechanisms involving Trusted Execution Environments (TEE)—as well as economic deterrents. Nodes are required to stake tokens, and failure to submit valid proofs during periodic challenges, or the discovery of other malicious activity, will result in the forfeiture of staked tokens. These incentives and penalties are designed to enforce honest behavior across the network.

Second, there is a potential risk of token inflation from a tokenomics perspective. Under CESS’s allocation model, a large proportion of tokens (up to 55%) is designated for node incentives. These tokens are gradually released into circulation over time, based on node contributions through mining rewards and service fee sharing. Although the total supply is capped at 10 billion CESS tokens, the annual release volume and its specific distribution curve have a direct impact on market supply and demand dynamics, as well as token value dilution. Compared to projects like Storj that may follow a relatively linear release model, CESS uses a contribution- and cycle-based dynamic release mechanism. Therefore, it is crucial to closely monitor the actual annual increase in circulating supply to assess any potential impact on token value.

Finally, the overall security of the network, especially against Sybil attacks or attempts to control a majority of the network’s computing/storage power, remains a critical concern. A common way to evaluate this threat is by estimating the economic cost for an attacker to control a certain percentage of network nodes. In the case of CESS, the cost of such an attack depends on the number of tokens an attacker must acquire and stake, as well as the computational resources and technical difficulty required to forge valid storage proofs. CESS strengthens resistance to such threats through its R²S consensus mechanism, which includes staking and credit scoring, the inherent complexity of PoIS and PoDR² proofs, and economic penalties for malicious behavior. However, as the network grows and token prices fluctuate, ongoing assessment and adjustment of attack costs are essential to ensuring long-term network security.

Conclusion

As the first decentralized data infrastructure with its own Layer 1 blockchain, CESS is transforming Web3 data storage and management through its innovative architecture, robust storage mechanisms, unique consensus algorithm, and multi-layered storage proofs. The platform’s versatility spans from basic storage services to AI training, digital asset markets, and user-friendly distributed network drives, showcasing its potential to reshape data valuation and circulation. Through a well-crafted tokenomics that incentivizes node contributions and network stability, CESS is building more than just a secure, efficient, and scalable decentralized storage network; it’s creating a foundation for data sovereignty, privacy protection, and ethical AI in the digital age. The project steadily advances toward its vision of a secure, transparent, and high-performance decentralized data value network.