Case Study: Constru accelerates AI experiments and cuts reproduction time with ClearML

Finding the Blueprints for Building More AI Experiments

Constru uses advanced computer vision to analyze construction projects, and its team needed a way to extend the platform’s AI capabilities without adding staff. As the number of experiments grew, they were spending too much time on debugging, documentation, experiment comparisons, and making sure results were reproducible. Constru chose ClearML to streamline its AI experiment management workflow.

With ClearML’s open-source MLOps platform, Constru automated experiment logging, simplified collaboration, improved experiment comparison, and made it easier to reproduce and deploy experiments across GPU clusters. The result was faster research and less manual overhead, with the team able to run up to seven times more concurrent experiments, cut reproduction time by an estimated 50%, and project about $1.3M in savings over 12 months.

Constru

Ron Zass

Chief Technology Officer