Case Study: Bentley Systems achieves faster AI development with Microsoft Azure Machine Learning

Bentley Systems helps close the global productivity gap with Azure Digital Twins

Bentley Systems, a software innovator serving infrastructure design, build, and operations, needed a faster and more repeatable way to turn paper-based and other “dark data” into usable digital information for its AI models. To address this challenge, Bentley Systems turned to Microsoft Azure, using Azure Machine Learning and PyTorch to formalize and accelerate its machine learning operations across its iTwin platform and other products.

Microsoft Azure helped Bentley Systems build a stable MLOps framework that lets data scientists work locally while automatically logging and moving code to the cloud for training and production. The result was a more reproducible workflow, higher developer satisfaction, and major speed gains: model development and delivery became three to five times faster, training time dropped from five days to one, and Bentley Systems was able to move six additional models into production quickly.

Bentley Systems

Pavan Emani

Vice President, iTwin Software Development