Skip to main content  
Implementing model and proof sharding for our
distributed verification system  
Enhancing consensus mechanisms to improve scalability 
Extending Symphony to
support aBFT protocols  
Enhancing network efficiency through data encoding and optimized message
distribution 
 
 
Open-sourcing tooling to
make it easier to run brokers 
Adding
scheduled and streaming transaction support  
Exploring posted price mechanisms for improved UX 
Battle-testing Resonance
alongside  the standard
EIP-1559 transaction fee mechanism 
Decentralizing  the Resonance
auctioneer 
Enabling cryptographic computational integrity and privacy gadgets for the
auctioneer via zkVMs & FHE 
 
 
Privacy Gadgets  
Developing novel privacy primitives (Cascade) that enables secure operations
on AI models while protecting sensitive data 
Incorporating existing Multi-Party Computation (MPC) and Fully Homomorphic
Encryption (FHE) gadgets for comprehensive privacy protection to the end user 
 
 
Agents  
Building the Agent Launchpad to enable easy deployment of autonomous agents
with built-in economic incentives and security guarantees 
Developing L2Beat-style analytics frontend to track agent performance,
revenue, and key metrics across different protocols 
Creating a robust agent framework that provides:
Standardized interfaces for agent-protocol interactions 
Built-in safety mechanisms and circuit breakers 
Composable agent primitives for rapid development 
 
 
Implementing agent-specific sidecars for efficient on-chain execution of
agent logic and state management 
 
 
Economic Design  
Adding support for restaking, allowing token holders to enhance network
security 
Continuing to reward meaningful network participation through
Proof-of-Useful-Work 
Implementing security measures to penalize malicious node behavior through
staking and slashing mechanisms 
 
 
Launching an auction system for model licensing rights 
Creating a dedicated marketplace for AI model intellectual property trading 
 
 
Asynchronous Execution  
Supporting
long-running, stateful, asynchronous precompile execution  
Adding the ability to
cancel ongoing computational tasks 
and Infernet operations 
Building more flexible transaction management capabilities 
 
 
Enhanced Infernet Node Management  
Updating our node software to support smarter work distribution 
Improving how nodes get compensated for their computational services 
Creating more efficient communication channels between nodes 
Integrating
Resonance as an Infernet node orchestrator  
 
 
Cross-Chain Compatibility  
Finalizing and securing our cross-chain communication contracts 
Adapting our technology to work with non-Ethereum blockchains 
Building comprehensive testing frameworks for cross-chain operations 
 
 
Expanded AI Capabilities  
Adding support for more AI model types, including image generation models 
Broadening the range of AI operations possible on the network 
 
 
Hardware Security Enhancement  
Enabling GPU-based Trusted Execution Environments for secure model operations 
Improving the security and efficiency of AI computations 
 
 
Additional Client Support  
Adding compatibility with more Ethereum execution clients, starting with Reth