1. The document discusses optimizing product delivery by analyzing data from the delivery lifecycle, including order placement, food preparation times, delivery routes, and external factors.
2. It notes challenges like inaccurate estimates of preparation times, traffic issues, and uneven runner availability that can negatively impact the customer experience. Data and machine learning models are proposed to gain insights and make more informed decisions.
3. Event-based data on orders will be collected including locations, times, customers, restaurants, runners and external conditions. This data along with collaborative and context-based machine learning models will be used to predict estimated times of delivery and optimize the process.