AI Agents for Decentralized Finance

Level: MA/MAP
Responsible person: Krzysztof Gogol
Keywords:  Solidity, Python, Blockchain, Smart contracts, Machine Learning

This thesis involves the development and training of an AI agent that curates liquidity pools in decentralized finance for selected crypto-tokens. Such AI agents can curate and maintain a pool at: 

• Decentralized exchange such as Uniswap v4 (with hooks) 
• Lending protocols such as Aave or Morpho 
• MEV - AI Agent that applies arbitrage or similar strategies to capture the market inefficiencies 

The goal of the AI Agent is to minimize costs for DeFi users and maximize returns for Liquidity Providers to DeFi protocols.  
 

References: 

Gogol et al. " SoK: Decentralized Finance (DeFi) -- Fundamentals, Taxonomy and Risks " (2023). 

40% Modelling, 50% Data Gathering and Analysis, 10% Documentation 

To apply for this thesis, please send your grade transcript to: Krzysztof Gogol (gogol@ifi.uzh.ch)