It was approximately the size and shape of a standard American shoebox. It had a display of ten small lamp lights labeled with the digits 0 through 9 and an attached microphone. Speaking the name of the digit into the microphone would cause the appropriate digit lamp to light.
Inside the box was a power supply, three analog audio filters and some (presumably) Diode-Resistor-Logic circuitry. The clever part of the design was in noticing that each digit name “Zero”, “One”, Two” … “Nine” had a front, middle, and ending sound. (Sometimes no middle). And that each sound was high pitched, middle pitched or low pitched. Example: “Five” is High-Middle-High. “Zero” is High-Middle-Low. The microphone was connected to the three audio filters for high, middle, and low pass. The filters latched the logic based decoder and switched one of the ten lamps.