We live in the era of Big Data. This phenomenon refers to the fact that the volumes of data around us are growing exponentially, the velocity of data generation and circulation is enormous, and also that the variety of data types and formats is wide and growing. On one hand, this phenomenon brings the prospect of a new level of data-driven decision-making and, on the other hand, completely new management and retrieval techniques are needed in order to keep pace with contemporary data. In this talk, we focus on three selected topics from the wide area of Big Data processing: 1) NoSQL databases: data stores that are more suitable for some Big Data problems then traditional relational databases; 2) deep convolutional neural networks: revolution in content-based multimedia retrieval; 3) large-scale similarity indexing and searching: latest achievements and practical examples.