An Automobile giant based in Australia was hit with a perfect storm of three major forces: challenging competitive segments, digital technological advances, and increasingly empowered consumers.

Advance their Sales processes by infusing them with AI-based intelligent prediction models. They had more than a decade of sales history data, and they wanted to mine through the data and extract hidden knowledge.

The following were the high-level features that are needed to be implemented in the application:-
• The automobile industry, like any other industry, has different market segments.
• It becomes essential to analyze your business intelligence and consider your competitor's intelligence and all other exogenous data that would impact the decision-making.
• To accommodate their new initiatives, they needed to have an intelligent and accurate sales forecasting system that significantly impacts their business.
• The prestigious autoregressive integrated models of moving averages(ARIMA) and the rarely implemented artificial neural networks (ANNs) and hybrid models were compared.
• RPA bots further utilized these predictions to check the inventory data to assess the needs and initiate appropriate spare parts order processing based on the vendor details and grades.

The solution utilized and connected the organizations' data silos, including the sales history, parts inventory, dealerships, CRM, and other exogenous data such as crude oil price variations, employment and unemployment data, and competitor intelligence data corresponding to the selected car segment for a selected geographical location. The solution process included cleansing of data to provide a neat orientation. The oriented data was then used to build various machine learning models. The machine learning models were built using Regression(XGBoostRegressor) for the car models whose standard deviation is more than a threshold and Time Series(N-Beats) for the car models whose standard deviation is less than a threshold.