The Cognitive Reasoner for Electronic Warfare Systems (CREWS) applies artificial intelligence signal processing to quickly classify complex, highly agile threat signals based on functional characteristics learned from the observed, possibly incomplete, waveform data. Research Associates of Syracuse (RAS) provides innovative signal processing solutions to challenging Electronic Warfare/Electronic Intelligence (EW/ELINT) problems. Initial targeted applications are Block 2 and 3 versions of AN/SLQ-32. CREWS technology is expected to reduce significantly (or even eliminate) reliance on threat libraries for emitter classification, which lowers maintenance cost and improves threat awareness in complex emitter environments. Phase I demonstrated accurate performance of selected machine learning classifiers trained using limited real radar data. Phase II is developing and demonstrating a full cognitive processing prototype using more extensive radar threat datasets. RAS seeks prime contractor support to integrate CREWS software into EW systems.