Roberto Esposito is an AI and Information Retrieval specialist currently working as a Research Engineer at Weaviate, a leading vector database company.

With a strong academic and research background, Roberto focuses on building scalable search solutions and machine learning systems. Before joining Weaviate, he conducted research on Approximate Nearest Neighbor (ANN) Search at the University of Pisa and the National Research Council (CNR) in Italy.

He holds a Master of Science in Computer Science with honors from the University of Pisa, where he developed deep expertise in algorithms, data structures, and artificial intelligence. He is also an open-source contributor, actively sharing his work on GitHub under the handle robbespo00.

As AI systems rely on semantic understanding, vector databases have emerged as a foundational technology for search, retrieval, and agents. This session explores how vector databases work under the hood, including the role of indexing strategies and compression techniques in balancing speed, accuracy, and cost. We will also compare single-vector and multi-vector approaches and discuss how each impacts retrieval quality in real-world applications. Through hands-on demonstrations of RAG pipelines and simple AI agents, we will translate theory into practice and show how vector infrastructure enables modern AI systems at scale.