Summary
Wilson Marcilio is a machine learning engineer and researcher with 11 years of experience building scalable ML systems, especially for embedding, manifold learning, conversational AI, and fraud detection. He has combined industry and academic rigor—moving from PhD research on visual analytics and dimensionality reduction to production roles at Globo, Nomic AI, and IBM—designing large-scale search and intelligent agents that handle millions of unstructured instances. Wilson excels at taking state-of-the-art transformer and encoder models from research to resilient cloud deployments, improving detection accuracy and conversational relevance in high-traffic environments. He brings deep expertise in embedding visualization and hierarchical summarization, a niche strengthened by international research collaborations (including Dalhousie) and FAPESP-funded projects. Now based in Paraná, Brazil, he continues to bridge research-driven methods with product-focused engineering as a Member of Technical Staff.
11 years of coding experience
5 years of employment as a software developer
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at Universidade Estadual Paulista Júlio de Mesquita Filho
Portuguese, English, French