Data Processing at Blue Yonder: One Supply Chain at a Time

Florian Jetter, Senior Data Scientist at Blue Yonder, joins Hugo Bowne-Anderson and James Bourbeau to discuss supply chain analytics at scale. 00:00 Going live! 02:30 What is Blue Yonder? 04:20 What is @Dask-dev? 05:45 How many clusters do Blue Yonder use? 08:30 Supply chain network breakdown 10:55 Breaking the problem of supply chain into pieces 13:15 Mid-size customer example 20:15 How did Covid-19 impact Blue Yonder's ML models? 23:11 Path to augmented dataset - How to get data into Dask 25:26 Connecting to Turbodbc 26:30 SQL query 29:10 Splitting data using some business logic 33:55 How not to kill the database 39:30 Persist data as a parquet dataset using kartothek 51:40 How to get more data out of your dataset 52:10 Enter: Machine learning 57:15 Where in the data process is Blue Yonder using Dask? 01:01:30 Wrapping up Thanks for tuning in! Check out Coiled Cloud for free today at https://cloud.coiled.io/.