We use a four-step process: first, collecting publicly available policy statements from official sources. Second, analyzing positions using natural language processing trained on historical policy data. Third, calculating alignment scores through vector similarity algorithms. Finally, presenting results in interactive visualizations with confidence indicators.

Candidates are added when they meet three criteria: (1) Have public policy statements in the FEC database, (2) Maintain an official campaign website, and (3) Pass our basic transparency requirements. Any user can request additions via our form.

Yes. We maintain complete political neutrality by: (1) Using open-source AI trained entirely on historical data, (2) Having no political donors or affiliations, and (3) Submitting to annual third-party audits. Our algorithm is open-source and publicly verifiable.

Our system polls official sources every 24 hours and immediately updates when candidates publish major events (debates, bill signings, or press briefings). Historical comparisons use the official records of those statements.