The village head election process that is currently underway in Tanjung Agung Village is still carried out conventionally, namely by punching the ballot paper on a piece of paper by showing the identity card and an election invitation letter to the committee. This conventional election has weaknesses, namely at the time of data collection there are double permanent voter data and some are not recorded, the process of calculating the results of the vote is slow, many ballots are damaged and there are still residents who abstain, especially residents who work outside the city. The system was built using the waterfall system development method with the programming language PHP, HTML, Notepad++ as a text editor and MySql as database storage, and Microsoft Azure Cognitive Services (Face Api) was used to create Face Recognition and was designed using the UML system development tool. The results of the black box testing conducted by involving 2 respondents stated that the functions in the system were running well and Face Recognition which was built using Microsoft Azure Cognitive Service (Face Api) successfully performed facial recognition with an average accuracy rate of 97% and the application which is built cannot store double data so that it can help reduce double permanent voter data during the data collection process by the committee.