Pydata CoBo #5: Virtual edition with REWE digital
Details
Hello PyData CoBo,
After our first successful virtual Meetup last month there will be another one in the coming week. Like last time, the plan is to have a single talk followed by a Q&A session.
After that there will be some break out sessions so that we can at least get the virtual equivalent of socialising and you are all welcome to stay a bit longer and chat about Data Science and beyond.
This time we are happy to welcome Sabrina Feder and Angelo Calci from REWE digital as our speakers. Their talk is titled "Forecasting E-Food sales to optimize automated warehousing at REWE digital" and you can find more details in their abstract below:
To cope with limited storage space for perishable food products and fluctuating customer requests, an accurate and stable approach to predict future sales is a crucial task in the E-Food business.
Reliably predicting purchase amounts poses a complex regression problem, due to the complicated and versatile nature of B2B and B2C E-Food Commerce, ranging from fast moving consumer goods to seasonal products or non-sellers.
We will introduce how the self-organised Scrum Teams at REWE digital are structured to support an efficient software development environment.
We continue by providing insights in our Apache-Spark based machine-learning approach to perform reliable and stable demand forecasts.
Moreover, we will show details of our scalable hybrid computing architecture on MapR as well as on Google Cloud and elaborate our deployment approach.
We are very much looking forward seeing many of you in this virtual format!
Your PyData CoBo Team
Ps: A link will be provided shorter to the date of the talk.