Inspiration

echnology has brought us incredible advancements—instant access to information, global connectivity—but it's also flooded the political landscape with misinformation and manipulative rhetoric. In the context of U.S. presidential debates, voters are left trying to sift through a sea of empty promises and confusing arguments. We were inspired by the need to empower citizens with clear, unbiased data, enabling them to make informed decisions in an era where trust in the political process is waning.

What it does

FairSight is a platform that analyzes U.S. presidential debate videos through voice and text classification. We use Gemini AI to evaluate the validity of arguments made by the candidates and generate data visualizations that highlight key insights. Our platform measures the impact of each speaker based on their characteristics and provides a user-friendly breakdown of the debates, making it easy for citizens to see through the noise and focus on what truly matters.

News bias detector Debates fact checking

How we built it

We used Whisper voice recognition software to identify speeches in live congressional broadcasts and classify statements by speaker, then integrated Gemini AI for argument validation. The platform uses data analysis to generate visual charts that showcase trends, speech impact, and argument strength.

With Ollama the news can be explained to us in a more simpler way, and the web application also provides friendly-learning resources.

Challenges we ran into

Another hurdle was integrating argument validity analysis in a way that could be easily understood by non-technical users, while maintaining the objectivity of the data.

Accomplishments that we're proud of

We are incredibly proud of having built a tool that can genuinely empower voters by providing clear, data-driven insights from the often confusing political debates.