LOFAR, the Low Frequency Array, is a new radio telescope constructed across Northern Europe which has recently started science operations. One of its key science goals is to detect and, where appropriate, respond to transient astronomical phenomena – such as supernovae and gamma ray bursts – as they happen. However, identifying these transients in the multi-terabyte LOFAR data stream is complex, and making the results easily accessible to astronomers is hard. I will describe the automatic transient identification systems we have put in place, and particularly emphasize how we use the high-performance, open-source MonetDB database system to make this possible.