Milo Coombs

Machine Learning Engineer @Moneycorp

London, GB

Innovating Financial Crime Prevention: Developing a new transaction monitoring platform within an innovation-focused team. This platform uses ML to enhance the speed and accuracy of detecting money laundering and terrorist financing.• Building a Multi-Agent System: Designing and implementing a multi-agent architecture where individual agents, each with a specialized ML model (e.g, Graph Neural Networks, Isolation Forests, Autoencoders), collaborate to reach a consensus on suspicious transactions.• Integrating Advanced AI: Utilizing Large Language Models (LLMs) to provide natural language summaries of agents\' findings and building a chatbot functionality to allow users to interact directly with transaction data.• Enabling Data Interaction with MCP: Leveraging the Model Context Protocol (MCP) to empower the chatbot to access and use internal tools, such as checking transaction histories and accessing Customer Due Diligence (CDD) information.

Durham University

Master's degree

Royal Grammar School Guildford

A-Levels