Hellas
Hellas is a decentralized serverless AI compute ecosystem secured by an L1 blockchain and powered by its own deep learning framework, catgrad. The network enables compute providers to join trustlessly and earn Hellas tokens while allowing users to run any model or custom code without centralized control. Its framework incorporates zero-knowledge verification for secure, privacy-preserving AI computations at scale.
- Artificial Intelligence
- Blockchain & Cryptocurrency
- Developer Tools
- Software Only
Funding
Funding not disclosed
- maven11.com (opens in a new tab)investor website
Founders
Product
Problem
Artificial intelligence is rapidly permeating every sector of society, yet its development and deployment remain concentrated in the hands of a few centralized entities. This centralization creates barriers to access, limits innovation, and concentrates control over transformative technology in ways that may not align with broad societal benefit. The current infrastructure for AI compute is also inefficient, opaque, and vulnerable to single points of failure.
Solution
Hellas provides a decentralized ecosystem for serverless AI compute, secured by an L1 blockchain and powered by its own deep learning framework called catgrad. The network allows compute providers to join trustlessly and earn Hellas tokens, creating an open, permissionless compute economy where anyone can participate. Users can run any off-the-shelf model or custom code on the network without barriers, centralized control, or reliance on traditional frameworks. The platform is engineered for distributed compute, enabling seamless scaling across decentralized networks for efficient AI training and deployment at any scale. Catgrad incorporates zero-knowledge proofs to ensure secure, verifiable AI computations while preserving data privacy and decentralized trust.
Target Audience
Primary customers are AI developers, machine learning researchers, and compute providers seeking decentralized, permissionless infrastructure for training and deploying models without centralized oversight.
Features
- Trustless compute provider onboarding with token-based incentives for a fully open, permissionless compute economy
- Permissionless execution of any off-the-shelf model or custom code without centralized control or framework dependencies
- Catgrad deep learning framework enhanced with zero-knowledge proofs for secure, verifiable AI computations
- Epic scalability engineered for distributed compute, enabling efficient AI training and deployment across decentralized networks
- Token-secured infrastructure with optimized, low-overhead compute and decentralized governance