Roney Lima do Nascimento is an applied generative-AI and LLM systems developer and a PhD candidate in Pure Mathematics at the University of São Paulo (IME-USP). He designs and deploys production AI systems spanning retrieval-augmented generation, agentic workflows, automated assessment, open-model fine-tuning and evaluation. His projects include an agentic engine with more than 121,000 lines of code and 3,800 automated tests, as well as an AI platform serving more than 15,000 monthly users. He is the author of Generative AI for Teachers, a Google Generative AI Leader, an IBM Generative AI Engineer, an invited keynote speaker and a peer reviewer for Springer. His current research focuses on trustworthy evaluation, provenance and hidden information leakage in agentic systems.