Article details
Two engineers at Revolut's crypto exchange developed a fully functional market-making system in just 30 minutes using Anthropic's Claude AI and the Model Context Protocol (MCP). The system, built as a side project, leverages AI to automate complex trading workflows—from portfolio screening to execution and monitoring—typically requiring weeks of development. The experiment highlights how AI agents can replace traditional trading platforms by executing strategies through simple text prompts, such as rebalancing a BTC/ETH portfolio based on market conditions. This raises critical questions about the future value of broker platforms in an AI-driven trading landscape. The integration of MCP allows AI models to interface with external tools like Revolut X's API without custom coding, enabling real-time decision-making. Analysts argue this could disrupt the $100 billion algorithmic trading market by democratizing access to advanced strategies. For traders, this means reduced reliance on proprietary platforms and faster deployment of adaptive strategies. However, challenges remain around regulatory compliance, data security, and the need for robust backtesting infrastructure. The implications for the crypto industry are profound. Revolut's experiment suggests a shift toward AI-native trading ecosystems where human traders act as supervisors rather than coders. Investors should monitor how competitors like Binance, Kraken, and FTX respond to this innovation. The adoption of open standards like MCP could accelerate AI integration across fintech, while regulatory bodies may need to update frameworks to address AI-driven market-making risks.
