Rafael Lacerda

I lead high-impact applied ML and GenAI initiatives that translate research into measurable product and business outcomes. My work focuses on shaping technical strategy and delivering end-to-end ML systems across recommendation systems, trust & safety, content intelligence, and large-scale LLM applications. I operate at the intersection of applied research, product strategy, and execution; owning problem framing, system design, experimentation and productionization. In recent years, I’ve driven initiatives including: - Embedding-based representations for content and users that power downstream classifiers and recommenders - LLM pipelines to extract novel metadata at a large scale. - Trust & Safety systems that reduce undesirable content and improve user experience - Cost and value-aware recommendation strategies balancing engagement, partner payouts, and business constraints - Internal tools and processes to supercharge developer productivity as well as stakeholder decision-making. I bring deep hands-on experience in NLP, information retrieval, learning-to-rank, recommendation systems, and applied deep learning. Today my focus is on technical leadership: setting direction, unblocking complex problems, and ensuring applied ML work is tightly aligned with product goals and real-world constraints.