Any research in the area of signal processing requires the development of large applications in a relatively short period of time. Unfortunately, research commonly suffers from a creative backlog due to rewriting of common functions and time spent debugging such things as file I/O. It would be ideal to have a large, hierarchical software environment which to support advanced research in signal processing. The ISIP Foundation Classes (IFCs) are being designed to meet this need, providing everything from complex data structures to an abstract file I/O interface. While the baseline public domain speech recognition system was written before IFC development, the next version will be distributed as part of a full programming environment.