Yes, I have done such a device. It did not display the speed on an LCD, but the lcd interfacing is very easy to do. The hard part is converting the timing of the pulses from the hall effect sensor into RPMs. The frequency is given by 1/t where t is the period of the pulses. Doing a "1-over" in micros is not always straightforward. Alternatively, you can count pulses in a given amount of time, but you need to have enough pulses in your sample period to give an accruate estimate of the speed (alternatively, you can digitally filter over the last n sample periods).