One of the tools most used by database administrators in the analysis of database systems quality of service problems are system event logs, which contain detailed information about the actions (SQL transactions) performed over time in a database system. Using these files, applying their experience and appealing to the expertize of data analysis tools, database administrators seek to establish event patterns that may identify the causes and reveal a path for establishing an appropriate solution to the problems. In this paper, we present and discuss a process mining system specifically designed and implemented to detect operational bottlenecks in relational database systems using transaction logs, providing expert information for their resolution, and contributing to improve system’s quality of service.