The following diagram outlines how the generated prediction is delivered to the user:
Step 1: Data inputs fed into prediction model.
Step 2: Prediction delivered to developers by the Prediction functionality of the API.
Step 3: Inetlligence delivered to end user.
Step 4: User provides feedback on accuracy of predictions.
The VQI API has two modes for functionalities. In Prediction mode, the API is used for pushing the predicted wait times to the application that is consuming the API. The Feedback mode is used for submitting crowd sourced data to Visa or the host application. These two functionalities can be used independent of each other.
The capability is grounded in implementing Queuing Theory with transactional data that Visa’s network sees and other external data sources to generate predicted wait times for merchants at a specific location. The API would take merchant name and location information from the main application and will send that information to Visa Queue Insight’s prediction model to generate a predicted wait time that can be sent back to the main application.