Demand Forecasting for Rental Car Offices

Problem

Stock planning is essential for car rental companies to meet their customer expectations and cost optimization. Manual fleet planning based on previous experiences is not meeting business and customer needs.

Action

Our solution included the following steps:

• Data preparation to obtain a structured data-set and to form a daily continuous data flow for automated demand planning
• Automatic outlier detection and clustering for segmentation
• Extensive feature engineering for demand signals
• Modelling and tuning for iterative evaluation
• User-friendly dashboard to monitor the forecast and key metrics
• Training the businesses for continuous model and process improvement

Tool Stack

• Azure, Python, PowerBI