Develop a low footprint, fast application that uses an available Speech Recognition API. Must be a Windows Application. The API must be from a vendor like Microsoft that has no royalties to use the API to develop an application.
The test application will have as input a list of words and phrases. This list will have a limited set of entries. Each entry in the list will be give an code. The application will be an "always on" application, listening to words being spoken. The job of the test application is to recognize the word from the set list and then display a log with the word/phrase + associated code. All other words should be ignored.
The application can be a single form. It needs to be able to work with multiple lists that will might from a text file. The text files will be in the format of one entry per line. Once loaded a list should be able to be quickly picked from a list of available lists. Example: one list might be a series of places (San Francisco, California) another list may be a list of Sport Teams like "Milwaukee Brewers". Multiple lists will allow quick testing. It might be best to save the data in a simple database once loaded. The list should be able to be edited once loaded.
Code should be written in such a way that an application can use the core functions. Prefer C# and .NET.