Klados

Klados allows to develop intelligent agents using our domain-specific language, which is then compiled into blockchain-compatible smart contract code.

With our platform, users can:

• Write sophisticated AI agent logic using our specialized code language
• Compile their agent specifications into optimized smart contract code
• Export deployment-ready code for direct implementation on blockchain networks
• Create autonomous, decentralized agents that execute complex logic within the secure environment of blockchain technology

Perfect for developers seeking to enhance their decentralized applications with intelligent, autonomous decision-making capabilities while maintaining the security and transparency benefits of blockchain infrastructure.

This project integrates several cutting-edge technologies to create a seamless bridge between AI agents and blockchain smart contracts. At the core of our compiler, we used ANTLR (ANother Tool for Language Recognition) to build a custom domain-specific language parser that translates our agent specifications into structured formats. For the secure execution environment, we implemented Marlin specifically for our Trusted Execution Environment (TEE), which ensures secure and private computation of agent logic while maintaining blockchain security guarantees. A critical component of our system is Recall, the first unstoppable intelligence network that empowers our agents to store, share, and trade knowledge onchain. This gives our agents persistent memory and the ability to collectively build on shared intelligence. The AI agent functionality is powered by AgentKit, which provides the foundation for our autonomous entity architecture. The backend is primarily built in Java, offering robust performance and extensive libraries that helped us implement the complex transformations needed to generate blockchain-compatible code.