Tudor Iustin Machine Learning Engineer & Donovan Agent Creator

Tudor Iustin is a Machine Learning Engineer with a focus on Large Language Models, applied AI, and agentic systems. He has developed the Donovan Agent, a self-learning computer agent that operates from the terminal and can take action across various systems, including files, commands, and browser flows. The Donovan Agent can turn successful workflows into reusable learned skills. Tudor Iustin's experience includes building a financial chatbot that simplifies complex financial reports into key business insights such as revenue, net income, and cash flow. He has also developed retail customer behaviour models and Python pipelines that improved targeting accuracy by 15-20% and reduced manual analysis time by 30%+. Additionally, he has created banking ML models for customer behaviour and risk detection, achieving improvements in model precision of up to 18% and reducing reporting preparation time by 20%. Tudor Iustin's technical expertise includes a range of tools and technologies such as PyTorch, Transformers, Keras, LangChain, and Kubernetes, reflecting his background as a Generative AI Engineer at BCG X and Machine Learning Engineer at BRD, as well as his experience as a Machine Learning Engineer Intern at Quantium.