Secure protocols
How can smart contracts and decentralized systems make their trust assumptions explicit—and expose failures before they cause harm?
Research & perspectives
Explore the thinking
behind the engineering.
Published work on intelligent agents, token engineering, and decentralized coordination—alongside the security and cryptography questions guiding our research.
6 articles
AI & agentsToken engineeringAI & agentsAI & agentsAI & agentsGovernance
Intelligent Liquidity Provisioning Framework
A comprehensive reinforcement learning approach to optimize liquidity provisioning on Uniswap V3 pools. This research explores how machine learning can enhance DeFi strategies.
The Blueprint for Designing Sustainable Decentralized Economies
Exploring the fundamental principles of token engineering and how to design sustainable token ecosystems that balance incentives, governance, and economic stability.
Intelligent Liquidity Provisioning Framework V2 — Implementation
The Intelligent Liquidity Provisioning (ILP) Framework V2 for Uniswap V3 pools is an advanced system that leverages agent-based modeling to optimize liquidity provision strategies.
ChatGPT vs TokenGPT — A Specialized Knowledge Assistant
A comparative analysis between general-purpose AI and domain-specific AI for token engineering. The article explores how specialized models can provide more targeted assistance.
TokenGPT — Your AI Token Engineering Assistant
An introduction to TokenGPT, a specialized AI assistant for token engineering that helps design, analyze, and optimize token-based economies using advanced language models.
From Tribes to DAOs — The Next Frontier in Social Governance
This article examines the evolution of governance systems from traditional tribal structures to Decentralized Autonomous Organizations (DAOs) and their impact on social coordination.
Research directions
Questions that shape our work.
Cryptography & verification
How do encryption, precise specifications, and formal reasoning support confidence in the systems people depend on?
Trustworthy AI systems
How can model evaluation, controlled tool access, and human oversight make AI applications more useful and accountable?