Lonneke Eijdems

Data Scientist

Lonneke, born in Limburg, studied in Groningen, lives now in Amsterdam. Loves sports, specifically hockey, race cycling and snowboarding.

Fun fact: I always try to go at least 3 weeks a year to Austra for snowboarding.

13:15 - 13:55

track Data

Online demand forecasting

Introductory session (Difficulty: Low)

During the ODFC session (formerly known as X12), we present the business questions that we solve and the impact that we are already making. We start by walking you through our quest in collecting and processing the right data. This required a scalable and robust data pipeline, which we built in DBT, but also a thorough understanding of the business and the data. Once you get a grasp of the data, we explain the modeling choices that we made, specifically the pros and cons of forecasting per zipcode (instead of on a higher aggregation level). Also, we present the details of the machine learning model that we implemented and how variables like promotions, holidays and minimal order value are used.