The Decentralised Finance (DeFi) community is continuously being affected by various types of security incidents that are causing high levels of financial loss. There have been multiple tools proposed for security attack detection, however, these cannot be used for general audit in DeFi applications as they focus on specific attacks or areas. This work aims to present how behavioural clustering and analysis can be leveraged to present insights into common user and smart contract behaviour in DeFi applications. Preliminary results are presented for two commonly known applications, Sushiswap and Uniswap V3 that lay down the foundation for general DeFi security audit.