Crescendo Speech is the first engine to support speaker independent speech recognition for large vocabularies. Available for both front and back-end use, the engine requires zero training with out-of-the box accuracy rates reaching over 95%.