Revisor is a neural network-based monitoring system designed for electoral oversight. It efficiently counts actual voter turnout and monitors compliance with electoral procedures using video recordings. The system employs advanced capabilities to learn and recognize physical objects, track movements, and detect voting events, achieving an accuracy rate of 98% under ideal conditions.

Users can deploy Revisor at scale, utilizing thousands of poll watchers across all polling stations in a constituency at a fraction of traditional observation costs. The tool is adaptable, allowing customers to train the neural network to operate with various voting procedures and electoral systems tailored to different countries.