Accelerating Large Scale Optimization Workflows with NVIDIA cuOpt and Metaflow
Unlock the power of open-source GPU-accelerated decision optimization with NVIDIA® cuOpt™ and Metaflow. cuOpt delivers high-performance optimization for complex decision problems — from vehicle routing to linear, mixed-integer and quadratic programs — at unprecedented scale and speed using GPU acceleration.
Together, cuOpt and Metaflow transform a high-performance solver into a complete, production-ready optimization platform. Metaflow provides a first-class integration with cuOpt, providing important benefits such as: GPU environments are available out of the box with no package setup, GPUs can be accessed cost-efficiently across clouds, and large numbers of optimization runs can be configured and orchestrated reliably in parallel.
Metaflow also enables end-to-end systems around cuOpt, integrating data access, observability, experiment tracking, and deployment into a single workflow — all securely running in your own environment.
In this webinar, we’ll walk through practical open-source examples that show how to use cuOpt and Metaflow together, from prototyping optimization models to running scalable, GPU-accelerated optimization pipelines in production.
Presenters
Burcin Bozkaya - Sr. Developer Relations Manager, NVIDIA
Eddie Mattia - Head of AI, Outerbounds
Who should attend
Data scientists and operations research practitioners building optimization models
Teams working on logistics, finance, or supply-chain optimization, fleet routing and scheduling
Organizations looking to productionize optimization at scale using open-source tools
Platform and infrastructure engineers responsible for GPU workloads and orchestration