Project Overview

The need for efficient, secure, and innovative AI solutions has never been greater in today’s rapidly advancing tech landscape. Traditional centralized AI infrastructures often struggle to keep pace with this demand and are typically limited by data privacy, security, and transparency concerns. Ritual is here to change that, offering a decentralized approach to reshape the AI space. By leveraging blockchain technology, Ritual tackles two major challenges that have long affected the industry: privacy and trust (About Ritual). This decentralized AI infrastructure was created to meet the urgent need for data processing solutions that safeguard user privacy without compromising security. Ritual’s system doesn’t just protect privacy; it also answers the growing call for transparency and trust in AI applications. With on-chain proof of AI interactions, Ritual creates a transparent ecosystem that enhances user trust (Ritual Document). The partnership with Nillion also introduces a trust-sensitive computing network, opening the door to groundbreaking AI applications across various sectors (Ritual Blog).

Background and Motivation

• The Rise of Decentralized Infrastructure

Decentralized AI infrastructure has gained traction due to the increasing demand for secure, efficient data processing systems. As AI becomes more embedded in different industries, there is a growing need for systems that handle sensitive data while maintaining privacy. Ritual, a Web3 platform, leads this movement by using blockchain to create a decentralized network for AI processing. This approach enables AI tasks to be executed while ensuring the confidentiality and security of the data (SiliconANGLE, Ritual FAQ).

• Addressing Privacy Concerns

Ritual was largely driven by the rising concerns over data privacy. Traditional AI systems typically require access to large datasets, often including personal, regulated, or proprietary information. This poses significant risks, as data breaches or unauthorized access could have serious consequences (Ritual Official Website).

• Building Trust and Transparency

Transparency is key when it comes to adopting AI technologies. Ritual offers on-chain proof of how AI models interact with data, providing a transparent record of data processing. This feature is particularly valuable in fields like healthcare, where handling sensitive financial and health data requires high trust. Ritual’s system increases credibility by offering clear audit trails, making it easier for users to trust AI applications (Ritual Official Website, SiliconANGLE).

• Improving Generative AI

Ritual’s infrastructure is also crucial in improving generative AI. It supports secure, private access to sensitive source data, essential for Retrieval-Augmented Generation (RAG) systems that use real-time data to enhance accuracy and reduce errors. This feature is particularly useful in industries where ensuring the accuracy and reliability of AI-generated content is critical (SiliconANGLE).

• Meeting Growing AI Demands

With the increasing demand for AI across industries, platforms like Ritual are becoming essential. As businesses look to leverage AI for a competitive edge, the need for secure, efficient, and transparent AI infrastructure becomes clearer. Ritual’s decentralized approach meets these needs and positions it as a key player in driving AI innovation within the digital economy (SiliconANGLE).

Core Technologies

Modular Execution Layer

Ritual’s architecture is based on a modular execution layer system called the Ritual Superchain. These layers are designed to handle different types of computational tasks, focusing on AI models. The execution layer is flexible enough to function as Layer 0, Layer 1, or Layer 2, depending on the need (Ritual FAQ). Each layer includes specialized state precompiles (SPCs), which are smart contracts optimized for specific AI functions like knowledge extraction, fine-tuning, and inference (Ritual Products).

• Decentralized Oracle Network (DON)

Ritual’s first phase, Infernet, is a decentralized oracle network (DON) that integrates AI into blockchain. It enables the deployment of AI models on any EVM-compatible chain, with plans to expand to non-EVM chains in the future (Ritual FAQ). Infernet bridges AI models and blockchain applications, allowing smart contracts to access AI models for advanced decision-making (What is Infernet?).

• AI Coprocessor

Ritual aims to become the AI coprocessor of the Web3 world. This involves creating a platform where AI models can seamlessly integrate with blockchain applications. The platform supports various AI models, such as large language (LLMs) and machine learning (ML) models, accessible through a universal API. This integration enables real-time AI-powered features like natural language interactions with smart contracts and automated risk management in lending protocols (Ritual Documentation).

• Security and Privacy

Ritual ensures security by leveraging Ethereum’s Layer 1 security via Eigenlayer as an initial security bridge. As the ecosystem grows, Ritual plans to develop its security measures. The platform uses cryptographic proofs to guarantee the privacy and integrity of computations, offering an extra layer of security for users and developers (Ritual Blog).

• Interoperability

Ritual’s General Message Passing (GMP) layer within the Superchain promotes interoperability between existing blockchains and the Ritual ecosystem. This enables Ritual to act as an AI coprocessor for multiple blockchains, increasing its utility and influence in decentralized environments (About Ritual Superchain).

• Nodes

Ritual’s infrastructure also consists of different types of nodes, each with its own functions and resource requirements. These include full nodes, validator nodes, proof nodes, model cache nodes, and privacy nodes, all contributing to the platform’s efficiency and stability. Ritual’s SDK provides developers the tools to quickly integrate AI into their decentralized applications (Ritual Documentation).

Infernet: Ritual’s First Product Stage

• Overview of Infernet

Infernet marks a significant advancement for Ritual, acting as a bridge between off-chain and on-chain computations. It is designed to integrate machine learning (ML) inference workloads with the blockchain, enhancing the capabilities of smart contracts. Infernet is a lightweight framework that allows developers to request off-chain computations from Infernet nodes and deliver the results to on-chain smart contracts via the Infernet SDK (Ritual Documentation).

Source: docs.ritual.net

• Architecture and Components

Infernet is built on an EVM-compatible framework, enabling seamless integration of AI/ML computations on the blockchain. This compatibility allows users to incorporate common AI workflows into blockchain operations. Infernet supports three types of jobs: off-chain (web2 requests), on-chain (web3 requests), and delegated (web2 and web3 requests). Jobs can be categorized into non-streaming and streaming. Non-streaming jobs follow a typical API request-response model, while streaming jobs offer real-time responses, useful for applications like chatbots (Ritual Documentation).

• Infernet SDK

The Infernet SDK is essential for developers subscribing to off-chain computational outputs. It provides interfaces like CallbackConsumer and SubscriptionConsumer, allowing smart contracts to use one-time or recurring computations. The SDK also supports asynchronous callbacks and manages payments between consumers and nodes (Ritual SDK Overview).

• Proof and Verification System

Infernet’s proof system is highly flexible, allowing the use of any proof system. Consumers assign validator contracts to verify the proofs provided by node operators. This flexibility ensures that proof systems can be tailored to the specific needs of various applications. Infernet’s contracts provide a framework for verifying computations, enabling consumers to request computation proofs from nodes, which are verified on-chain (Ritual Documentation).

• Payment System

Infernet 1.0.0 introduces an on-chain payment system compensating node operators for their computational work. Developers pay for their computations, and both parties can register on-chain Infernet wallets to facilitate transactions. This payment system is integral to Infernet’s economic model, ensuring that node operators are incentivized to provide computational resources (Ritual Documentation).

• Use Cases

Infernet supports a range of applications, particularly those involving ML inference workloads. For example, developers can delegate intensive tasks like ML inference or ZK proof generation to off-chain nodes and use the output in smart contracts through on-chain callbacks. This is especially useful for applications requiring real-time data processing and decision-making (Ritual SDK Overview).

• Real-World Implementation

Infernet has been applied in various scenarios, demonstrating its versatility and effectiveness. For instance, developers can add new traits to NFTs based on user input, enhancing NFT collections. This process involves off-chain stable diffusion workloads that analyze NFT images, add new traits, and return the updated images to smart contracts. Infernet can also run automated risk models for DAOs, providing quantitative risk scores for proposals based on on-chain parameters (Ritual SDK Overview).

• Future Prospects

As Ritual’s first production implementation, Infernet sets the stage for the future development of the Ritual network. Connecting off-chain computations to on-chain smart contracts opens new possibilities for AI and blockchain integration. Continued development of Infernet and its tools, such as the Infernet SDK, could lead to further innovation in decentralized AI infrastructure (Ritual Products).

• Summary

Infernet represents a crucial evolution in the Ritual network, offering a comprehensive framework for integrating AI/ML workloads into blockchain environments. Its architecture, components, and features make it a powerful tool for developers seeking to enhance smart contracts with advanced computational capabilities. As Infernet continues to evolve, it will play a key role in shaping the future of open AI infrastructure on the blockchain (Ritual Documentation).

Ritual Superchain: The Second Product Stage

Ritual’s next product, the Ritual Superchain, will be a customized execution layer built to support AI-native operations, empowering a new wave of applications at the intersection of crypto and AI (Ritual Introduction). It is expected to launch in the coming months, and the product architecture is illustrated in the diagram below (Ritual Blog).

Source: ritual.net

Ecosystem Partnerships

• Partnering with EigenLayer

Some key features of Ritual require strict economic and security assurances. EigenLayer’s restaking mechanism allows Ethereum stakers to extend the strong economic security of Ethereum to other applications in exchange for diversified income streams. By combining the two, Ritual can leverage the properties inherited through EigenLayer to build new Active Validation Services (AVS) for various AI workflows while ensuring economic and security guarantees. EigenLayer enables Ritual to deliver strong decentralization and security for AI operations from day one while also offering EigenLayer operators a new way to earn yields (Ritual Blog).

• Partnering with Story Protocol

Ritual’s partnership with Story Protocol makes AI models and their outputs traceable and monetizable. Together, they can create a massive intellectual property graph, allowing developers to track how their AI models are deployed and fine-tuned. They also ensure that AI-generated content is watermarked and tracked on-chain. This partnership is a fresh effort to merge blockchain technology with the profitable field of AI development (blockworks).

• Partnering with MyShell

Ritual has formed a strategic partnership with MyShell, a platform that uses Ritual’s infrastructure to support a variety of open-source AI models and applications. This partnership highlights how Ritual’s infrastructure can enable new features and functionalities previously unavailable in traditional AI stacks. Integrating MyShell into the Ritual ecosystem is expected to pave the way for a new generation of native applications (Ritual Blog).

• Partnering with Celestia

Celestia has partnered with the decentralized AI computing platform Ritual to integrate its modular data availability (DA) network with the Ritual Chain. The partnership aims to leverage Celestia’s DA capabilities to improve the efficiency and reliability of Ritual’s AI operations. Ritual users can publish data generated in their workflows as blobs onto Celestia’s DA layer, which includes Ritual Chain (also known as the Ritual Superchain) and the decentralized oracle network Infernet (Ritual Blog).

Celestia’s modular design helps reduce transaction fees by minimizing network congestion. It is crucial for Ritual as it allows AI models and services to be deployed at affordable prices, making them accessible to a wider audience (SiliconANGLE).

This partnership is strategically important for both parties. For Celestia, it’s an opportunity to showcase its modular architecture’s capabilities in real-world applications, potentially attracting more developers and projects to its platform. For Ritual, integrating with Celestia strengthens its ability to deliver decentralized AI services, positioning it as a leader in the emerging AI space on the blockchain.

• Partnering with Nillion

Ritual’s collaboration with Nillion, a trust-sensitive computing network, is set to revolutionize the handling of sensitive data in AI inference processes. By integrating Nillion’s “blind” computing technology, Ritual allows developers to maintain the confidentiality of both user input data and AI models. This opens up new possibilities for innovative AI applications, enabling developers to explore more complex and sensitive use cases without compromising data security (Ritual Blog).

The partnership between Ritual and Nillion also paves the way for a wide range of use cases beyond traditional AI applications. The platform could be used for IoT processing, where secure data handling is critical, and has potential in areas like price prediction, where protecting proprietary models and user data is essential. This system offers a secure alternative to current anonymization solutions, further enhancing its applicability across various sectors (SiliconANGLE).

Core Team

Source: rootdata.com

• Niraj Pant: Co-founder of Ritual. Previously, he was a general partner at Polychain Capital, where he invested in early-stage cryptocurrency projects. He also conducted privacy research at the Decentralized Systems Lab at the University of Illinois Urbana-Champaign and co-founded Source Networks as CTO. Niraj holds a degree in computer science from the University of Illinois Urbana-Champaign.
• Akilesh Potti: Co-founder of Ritual. A former partner at Polychain, Akilesh has experience in machine learning at Palantir, high-frequency trading at Goldman Sachs, and research in machine learning at MIT and Cornell University.
• Anish Agnihotri: Founding member of Ritual and an independent researcher of MEV. He has worked at Paradigm and Polychain, where he built high-performance trading systems, and also worked on automation testing at 1Password.
• Eva Zhang: Founding member of Ritual and a venture investor at Sequoia Capital, as well as the founder of Socket. She graduated from Stanford University.
• Igor Sylvester: A founding member of Ritual, previously an engineering partner at Trust Machines. He holds a degree in computer science from MIT.
• Saneel Sreeni: Founding member of Ritual and a venture partner at Accomplice, where he invests in and mentors ambitious startups like Praxis, Kaleidoscope, Veera, Eclipse, and Karate Combat.
• Emperor: Researcher at Ritual.

Ritual is led by co-founders Niraj Pant and Akilesh Potti, who have extensive experience in blockchain and AI technologies. The team is supported by a strong advisory group, including Illia Polosukhin, co-founder of NEAR, and Arthur Hayes, co-founder of BitMEX. This leadership and advisory team provides strategic insights to ensure Ritual’s continuous success in the decentralized AI space (Ritual Blog).

Funding Situation

Source: ritual.net

In November 2023, Ritual raised $25 million in Series A funding led by Archetype. Other key investors included Accomplice, Robot Ventures, dao5, Accel, Dialectic, Anagram, Avra, and Hypersphere. The funding will expand the team, grow the developer network, and provide seed capital for network growth, helping propel Ritual into its next stage of innovation and progress.

Roadmap

Although Ritual’s roadmap hasn’t been released yet, it’s expected to outline the platform’s future developments and key milestones. The team is committed to building transparently and regularly shares updates through social media and the product page on their website (Ritual FAQ). The next phase of Ritual’s development will focus on creating its own sovereign chain (Ritual Superchain), complete with a custom virtual machine (VM), which will act as a coprocessor, further enhancing its functionality and scalability.

Conclusion

As Crypto and AI evolve, Ritual emerges as a standout innovator in decentralized AI infrastructure. By addressing critical challenges like privacy, transparency, and security, Ritual is well-positioned to transform how AI integrates across industries. Its strategic partnerships with Nillion and Celestia extend its capabilities further, providing scalable and secure AI solutions that are cost-effective and reliable (SiliconANGLE). Ritual’s modular execution layer and decentralized oracle network represent a significant step forward in bringing AI to the blockchain, increasing the practicality and reach of blockchain applications. The $25 million raised in Series A funding demonstrates strong investor confidence in Ritual’s vision, enabling the platform to expand its developer network and innovate (Cointelegraph). With the launch of Ritual’s sovereign chain in the coming months and the ongoing expansion of its ecosystem partnerships, Ritual is set to play a pivotal role in democratizing AI and building a more secure and innovative crypto economy for the future (Coindesk).