Data journalism and digital media organization
Detection of missing data in election tally sheets
We implemented an AI-powered solution combining computer vision and robotic process automation (RPA) to analyze more than 130,000 election records, automatically identifying potential inconsistencies and streamlining large-scale document analysis.
- Industry
- Capabilities
- Regions
- 🇦🇷 Argentina
- Published
- July 2026
The Challenge
To shed light on an issue present in every election and encourage discussion about Argentina's voting system—whose primary voting instrument, the party-list paper ballot, has shown several shortcomings—our client sought to analyze the validity of manually completed election tally sheets.
The Solution
Using robotic process automation (RPA) and computer vision techniques, we analyzed approximately 130,000 tally sheets from Argentina's 2021 primary elections. Through a computer vision algorithm specifically designed for this purpose, the solution examined the documents and identified potential errors in their completion with a very high level of accuracy.
Benefits
- Analysis of multiple data sources.
- Automated download, analysis, and reporting processes.
- Fast and intuitive visualization of the resulting statistics.
- SIGMA Awards recognition for the world's best data journalism team.