Blending Optimization

PBLeiner has a warehouse with hundreds of bags of gelatin, that can be combined in a blending process that will result in a ready-to-ship produce for the client. We designed

  • a ML model that predicts the quality of each blend and
  • an algorithm that aids the blending manager in picking the right bags from the warehouse to minimize overdelivering quality to the customer, to minimize stock loss and to improve warehouse efficiency.

Moreover, we built a Power Automate flow as the connection between a custom-built Power App and the models, which reside in a Databricks workspace.