Kensington, England, United Kingdom
Machine Learning Engineer with experience in fraud detection, predictive modelling, LLM applications, AI agents, RAG, data engineering, and full-stack development. Skilled in building and deploying machine learning solutions, automating workflows, processing large-scale datasets, and enhancing business systems using Python, PySpark, Django, Docker, Kubernetes, and data pipeline automation.
• Develop machine learning models to evaluate existing fraud detection rules and identify new rule opportunities. • Apply machine learning and deep learning techniques to improve fraud risk scoring and overall detection performance. • Build LLM- and agent-based solutions to generate fraud alert explanations and automate fraud detection workflows. • Develop AI agents and RAG-based solutions to automate internal tasks and improve access to enterprise knowledge. • Containerize applications with Docker and deploy them on Kubernetes to support scalable, reliable operations. • Process large-scale datasets using PySpark to support fraud analytics and model de
Built machine learning models to predict customer purchasing behaviour for personalized marketing and to predict the upcoming auction results, resulted in reducing marketing cost by approx. 15%. Analysed complex datasets to gain insights on customer preferences and generating reports using Python and R; Maintained and enhanced CRM system to meet new business requirements and integrated multiple existing systems using Python and PHP; Managed a data migration project from multiple data sources and built data pipeline to automate data transfer using Python and Bash script;
Analysed complex datasets to gather insights on customer collecting interests and purchasing patterns, thereby enhancing customer service and guiding management decisions; Developed new modules on CRM system using Python and PHP to support new business goals and ongoing data transfer between systems; Managed a data migration project to a new auction management system, which involves a huge volume of data cleaning and processing; Automated data transferring pipeline and daily tasks using Python and Bash script;
Independently developed a complex web-based management system using PHP/Zend and MySQL, functions include live auctioning, lotting, invoicing, stock management, shipping management and pdf auction catalogue creation; Analysed complex datasets to extract customers collecting interests and purchasing behaviour to set up catalogue mailing list, successfully identified VIP and inactive clients, which reduced marketing cost; Automated several general administration tasks using Python and Linux Bash script.