DataGEMS at the 27th IEEE International Conference on Mobile Data Management

A thought-provoking question was explored by Dr. Georgia Koutrika, Research Director at Athena Research Center and Coordinator of the DataGEMS project, in her invited talk at IEEE MDM 2026 in Athens (June 29th – July 2nd, 2026).

The presentation explored how Artificial Intelligence is beginning to reshape the core of database management systems. It discussed emerging approaches such as learned query optimizers, which leverage past executions and data characteristics to improve query planning, and LLM-based semantic operators that allow systems to interpret, align, and reason over heterogeneous data, schemas, and modalities.

It also addressed key challenges in the field, including the difficulty of ensuring that learned optimizers generalize effectively and behave consistently across different workloads, the need to define semantic operators with the formal properties required for systematic optimization, and the challenge of establishing reliable evaluation methodologies for comparing emerging database architectures.

These challenges strongly resonate with the vision of DataGEMS, which aims to enable intelligent discovery, integration, and exploration of complex and heterogeneous datasets through advanced data management, AI, and machine learning technologies.