This is a keyword spotting system that searches for keywords in a written transcription of audio data (in Swahili) produced by a 1-best speech recognition system. It was coded in the context of the University of Cambridge's MPhil in Machine Learning and Machine Intelligence ( https://www.mlmi.eng.cam.ac.uk/ ). The data came from the Babel Project, and is not provided with the code. A more complete description of the code, as well as of some experiments run using it, can be found in "KWS_Swahili_report.pdf".