loading

Business Impact

  • 98%

    Prediction accuracy

  • Sub-second response time

  • 5.3M

    Predictions made over 10k bills

Customer Key Facts

  • About : Award-winning publishing company
  • Size : 300+ employees
  • Location : Washington, D.C.
  • Industry : News & Publishing

Problem Context

Congressional Quarterly Roll Call produces a number of publications that report primarily on the United States Congress. Their existing solution for bill prediction had a limited number of parameters; thus, they wanted more parameters which could better predict the probability of a legislative bill passing the US Congress to help their customers in government relations make strategic decisions.

Challenges

 

  • Consideration of hundreds of variables from disparate sources to generate a useful and most significant prediction
  • Daily execution of the model on a large volume of bills
  • Maintaining a track of amendments proposed on a bill to accurately predict the outcome

Technologies Used

Amazon EC2
Amazon S3
AWS Lambda
Amazon API Gateway

Predicting the Passage of a Legislative Bill in Both Houses of U.S. Congress with Predictive Analytics and Regression Modeling

Solution

Quantiphi achieved impactful results by identifying the most influential features in predicting a bill as passing or failing. The probability prediction solution was also integrated with the customer’s website to automate the result generation and keep the users updated.

Results

  • Achieved an accuracy of over 98 percent for predicting the outcome of a bill
  • Displayed the top factors associated with the probability of a bill passing or failing
  • Designed subsequent API calls to connect the client’s front end to the bill scoring

Looking for similar project?

Let's Talk

Explore more

Get your digital transformation started

Let's Talk