The talk at the FOSDEM conference in 2023 focused on graph aggregation and grouping in streaming graphs. The speaker explained the concept of graph grouping, which allows for the grouping of graphs based on certain attributes, such as time, structure, and content. This enables the execution of graph analytics on group graphs, providing deeper insights and allowing for various applications, such as pre- and post-processing steps in graph analysis, understanding the graph stream in more detail, and revealing hidden information. The speaker also introduced the Graph Stream Zoomer project, which provides a Java API for graph grouping. They discussed the challenges faced in implementing the project, such as finding the optimal graph representation and ensuring scalability. The speaker also shared examples of how to use the grouping algorithm to group graphs based on different criteria and perform aggregate functions on the grouped graphs. They mentioned that the project is still under development and encouraged contributions. Finally, the speaker mentioned related projects, Gradube and Temporary Graph Explorer, and provided links to their GitHub repository.