Dirk Schumacher

Decision scientist
Dirk has worked in various software development and data science roles over the last 10 years. At university he first discovered operations research and has been fascinated by it ever since. He enjoys swimming and contributes to open-source software, often at the intersection of R and mathematical optimization.
Release: Shadow testing for decision models in Nextmv

Shadow testing isn’t just for the MLOps crowd. Decision algorithm teams can have it too for their optimization models solving routing, scheduling, allocation, and fulfillment problems across supply chain, retail, healthcare, and more.

Solve a vehicle routing problem (VRP) using R and Nextmv Cloud

In a few simple steps, you can formulate, solve, and visualize a capacitated vehicle routing problem with the R programming language.

Switchback testing decision models: Demo using Nextmv

Use Nextmv to compare two decision models operating in production while accounting for network effects

What is switchback testing for decision models?

Switchback testing is a key part of the DecisionOps workflow. Get a quick primer on what it is and when to use it in this 2-minute video.

Comparing decision models in operational environments with switchback testing

When you’re ready to have a candidate model make true operational decisions, it’s time for switchback testing. Kick off an experiment and analyze how your new decision model measures up to your current model in production.

Determining decision model readiness using shadow tests and acceptance tests

How do you feel about the decision model updates you ship to production? Acceptance and shadow testing are two ways to gain confidence across model performance for business KPIs and stability indicators. We’ll show you how.