UiPath re-architected its infrastructure to support high-scale intelligent document processing using a shared Google Cloud GPU fleet powered by AI Hypercomputer and Dynamic Workload Scheduler.

Key Details

• UiPath built a shared GPU fleet managed by its Machine Learning Services platform on Google Cloud
• Balanced A3 VM instances with NVIDIA H100 GPUs for training and G4 VM instances with NVIDIA RTX Pro 6000 GPUs for inference
• Utilized Dynamic Workload Scheduler to book compute capacity in advance and stabilize costs

In This Story