Building an EVM-Compatible Decentralized Storage Platform for Data-Heavy Workloads

Snapshot

  • Focus: Designing and implementing an EVM-compatible decentralized storage platform for verifiable, large-scale data
  • Context: Blockchain applications lack native, scalable storage aligned with Ethereum security and developer workflows
  • Who it’s for: Moonbeam Foundation and developers building data-intensive and AI-driven applications
  • Outcome: A storage stack combining a StorageHub-based protocol, an EVM-compatible execution layer, and a security model connected to Ethereum via restaking.

Context

While smart contracts and execution environments evolved rapidly, developers continued to rely on centralized or semi-trusted systems for data storage. This introduced architectural fragmentation, weakened trust guarantees, and created a poor developer experience.

At the same time, emerging AI-driven workloads introduced new requirements. Agent memory, datasets, and verifiable logs demand strong guarantees around integrity, availability, privacy, and audit-ability. Existing decentralized storage solutions often required separate networks, bridges, and trust assumptions, making them difficult to integrate cleanly into application architectures.

Moonbeam required decentralized storage to function as a first-class system aligned with Ethereum execution, security, and developer tooling. Moonsong Labs supported the design and implementation of DataHaven to meet those requirements for data-heavy, real-world applications.

The Challenge

The core challenge was not simply to provide decentralized storage, but to do so in a way that reduced architectural complexity while remaining compatible with existing developer ecosystems.

Key constraints included:

  • The absence of native, scalable storage primitives in blockchain execution environments
  • Fragmented architectures caused by stitching together multiple chains, bridges, and storage networks
  • The need for Ethereum-compatible security guarantees rather than introducing a new, isolated trust domain
  • Requirements for verifiable, auditable, and privacy-preserving data suitable for AI-era workloads
  • A strategic desire to extend beyond Polkadot while preserving Substrate-based system strengths

The goal was to design a storage platform that unified execution, storage, and security into a coherent system without weakening trust assumptions.

Our Approach

Moonsong Labs approached DataHaven as a system-level infrastructure project rather than a standalone application.

Key elements of the approach included:

  • Architectural foundation: Building on the work established in StorageHub.
  • Execution and storage co-design: Embedding storage directly into an EVM-compatible Substrate chain rather than treating it as an external dependency
  • Security model: Securing the network via restaking mechanisms that derive economic security from Ethereum, rather than launching a new validator set.
  • Developer-first integration: Treating SDKs, documentation, devnets, and workflows as core deliverables, not ancillary tooling
  • AI-aware architecture: Designing storage semantics around verifiability, encryption, and audit-ability as baseline requirements

Rather than optimizing for a single application, architectural decisions prioritized generality, composability, and long-term extensibility.

Execution

DataHaven was implemented as an integrated decentralized storage platform composed of multiple tightly coordinated components.

At the protocol layer, the system leveraged a StorageHub-based decentralized storage protocol, providing dedicated storage provider roles, cryptographic integrity guarantees, and a data model optimized for large-scale storage.

At the execution layer, the DataHaven chain was implemented as an EVM-compatible Substrate blockchain. Storage primitives were embedded directly into the execution environment, reducing fragmentation and enabling smart contracts to interact natively with stored data.

Security was aligned with Ethereum through integration with EigenLayer via restaking, rather than introducing a separate validator security model. Interoperability between Ethereum and DataHaven was implemented through trust-minimized cross-chain communication, using a dedicated relayer and Solidity-based contracts to exchange messages without relying on centralized bridges.

The developer surface included SDKs, documentation, local development environments, and end-to-end workflows for uploading, retrieving, and verifying data. These components were developed alongside the protocol to ensure that the system could be exercised as a complete product rather than an isolated protocol.

Results

Outcomes from the DataHaven project are best evaluated through observable, verifiable artifacts and network operation milestones.

  • An integrated decentralized storage platform: DataHaven combines a StorageHub-based protocol, an EVM-compatible execution layer, and a unified security model into a single, cohesive system.
  • Ethereum-derived security model: Validator security is derived from Ethereum economic guarantees via EigenLayer restaking rather than a bespoke or standalone security mechanism.
  • Operational testnet launch: The DataHaven testnet was launched on December 17, 2025, providing a live environment for exercising storage, execution, and interoperability workflows.
  • Open-source implementation: Core protocol, relayer, and supporting components are publicly available, providing an auditable record of architectural and engineering decisions.
  • Clear path to production deployment: The system has progressed from protocol design through testnet operation, with mainnet deployment planned as the next phase.

These outcomes are best represented through public repositories, network launches, and ongoing protocol development rather than claims that may change over time.

Closing Reflection

DataHaven illustrates how decentralized storage becomes more tractable when storage, execution, and security are designed as a single system rather than independent layers. By deriving security from Ethereum, embedding storage directly into the execution environment, and treating developer experience as a core consideration, the project reduces architectural fragmentation for data-intensive applications.

The work contributes to ongoing efforts in Web3 infrastructure to design cohesive, system-level architectures capable of supporting real-world, data-heavy, and AI-driven workloads.

If you’d like to review the implementation and supporting materials:

Additional progress and network updates are publicly visible via the DataHaven testnet announcement and ongoing development activity.