Shashank, a distributed middleware researcher from the University of Toronto, is developing Ekai to serve as a universal context layer for agentic AI. This middleware addresses the issue of vendor lock-in found in tools like Claude Code or Cursor by allowing developers to switch between different inference providers without losing their memory or preferences. By integrating a Web3 ownership model, Ekai enables users to maintain sovereignty over their data through a verifiable cloud infrastructure, potentially utilizing Trusted Execution Environments (TEEs). The system simplifies the developer experience by removing the need to manually manage local vector or graph databases. It functions as a local gateway that standardizes API endpoints, similar to OpenRouter, while incorporating a wallet for X402 payments to facilitate seamless access to various models without requiring individual API keys.
Sign in to continue reading, translating and more.
Continue
