This webinar, organized as part of the EXA4MIND project, focuses on the fundamentals and practical use of vector databases for similarity search tasks in AI workflows.

The session begins with an introduction to core concepts such as vector embeddings and how they enable semantic understanding of data. Different types of vector search—including image vector search, text vector search, and hybrid approaches—are discussed conceptually to provide foundational understanding.

In the hands-on part of the webinar, Milvus is used to create a vector database collection and load both image and text embeddings. Participants are guided through code demonstrations showing how to perform various search operations and analyze the results in real-time.

This practical walkthrough aims to illustrate how vector search capabilities can be integrated into AI-driven systems to support advanced analytics and retrieval tasks at scale.