Introduction
Waternet is the only water company in the Netherlands that manages the entire water cycle. They supply drinking water in Amsterdam and its surroundings. In the area of the Amstel Gooi and Vecht Water Board, they work on strong dikes and clean water.
Customer
Waternet
Tools/Techniques
Azure (DevOps), Bicep, Python
Services
Data Science & AI, ML Engineering
Industry
Utilities
The goal-driven data solution
The Waternet Datalab team aims to execute projects that are actually implemented. A robust MLOps workflow is essential for this. As part of the Datalab team, we have set up the MLOps workflow and maintain it. This includes an MLFlow model registry, CI/CD templates, and the implementation of code and models into production. We have developed an ML model to minimize nitrous oxide emissions at the wastewater treatment plant. Currently in development is a model to limit the use of iron chloride in the pre-treatment of river water for drinking water production.
“It feels good to develop and sustain solutions with a direct societal impact. All of the team’s projects have the focus to go to production, which makes the ML Engineering work even more interesting.”
Check out other case studies,
