Booster is divided into four sections to continuously optimise results and improve efficiency. These steps include data collection, data preparation, model development, and data iteration. The cycle of monitoring and evaluating results allows for continuous updates and improvement.

Data Collection

The initial stage is data collection, which comprises of extracting, transforming, and loading data from the Production Database and ADVANCE’s Database after collecting from multiple sources.

Data Preparation

Data preparation is the next step- ensuring that the data is in a format recognisable by the engine. The data is cleaned and matched against features for compatibility, and then samples are arbitrarily chosen and vetted prior to the actual processing of the entire data population.

Model Development

Upon completion of data validation, the data is sent through the model to produce the most optimal outcome. Data Analysis and Binning are used in conjunction with Machine Learning techniques for the model’s ‘Intelligence Profiling’ of customers.

Data Iteration

The final stage of the loop is data iteration, which is the the overall monitoring and tuning of the system. This allows Booster to constantly re-train its model and employ techniques to develop a better product.