(by COSMOTE – Konstantinos Filis)

The TRACY project faced the challenge of accessing, in an automated way and without violating subscriber privacy regulations, non-content mobile telecommunications data, both subscriber and network-related, including the subscriber number, the type, date, time and duration of a call, as well as the user location (denoted by the area covered by the antenna the user is connected to), owned by the Mobile Network Operator (MNO).  Currently, these data can be obtained only through special requests from law authorities, due to legal restrictions, resulting in significant delays that make the swift identification of crime suspects impossible. To meet the project objectives, that is to identify the suspects, more victims or potential perpetrators, at the early stages of the investigation, AI-based algorithms were developed which were validated through specific use cases / scenarios utilizing synthetic data.