Revisor

What is Revisor?

Revisor is a software package that uses neural networks to monitor how elections are run. It acts as a digital watchdog for voting processes: it counts the real number of people who vote, spots rule-breaking, and speeds up slow manual recounts. The system is built on computer vision and trainable neural networks that recognize physical objects, track movements, and distinguish actual voting events at polling stations. Revisor counts voters autonomously and is reported to be up to 98% accurate at finding discrepancies in voter turnout. It can be customized for different voting procedures and electoral systems, making it adaptable to elections in many countries.

Who created Revisor?

Revisor was developed and launched by Poll Watcher, an AI company specializing in neural network-based software for monitoring electoral procedures.

What is Revisor used for?

• Detect ballot boxes in video recordings.
• Count the number of voters who cast ballots.
• Identify polling stations where turnout may have been falsified.
• Detect multiple types of violations.
• Draft formal complaints when issues are found.
• Speed up manual recounts.
• Process millions of hours of video.
• Train on different electoral systems, making it adaptable to any country regardless of its election laws.

Who is Revisor for?

• Journalists reporting on election integrity.
• Election observers monitoring the process.
• Investigative teams verifying electoral irregularities.

How does Revisor work?

Revisor’s neural networks are trained to recognize objects, track movements, and identify voting events in video footage. In practice, the system detects and outlines ballot boxes in recordings, counts voters, flags falsified turnout, identifies violations, and helps draft formal complaints. Because it processes large volumes of video, it can cover substantial amounts of polling station footage without manual review.