How to ship better plans: Decision parameter testing and tuning in practice
How confident are you that your model's parameters are actually right? Most teams tune by gut: change a number, run the model, eyeball the output, repeat. It works until it doesn't, and when it doesn't, you get plans that are mathematically optimal but operationally broken.
How to set up CI/CD with GitHub Actions for your Nextmv decision app
This video tutorial demonstrates how to implement a continuous integration and continuous deployment (CI/CD) decision workflow using Nextmv Cloud and GitHub Actions. The process begins with setting up the Nextmv CLI and initializing a sample vehicle routing application, known as a community app, which is run locally to verify basic functionality.
Decision systems in mobility and commerce: In conversation with Goda Ramkumar
Goda Ramkumar, Vice President of Data Science at Swiggy, sat down with Carolyn Mooney, Nextmv CEO, to discuss the evolution of their decision systems in mobility and commerce, from strategic network design to real-time assignments for delivery partners and orders. The conversation covers the journey from greedy algorithms to more sophisticated mathematical optimization approaches.
Exploring new frontiers in decision optimization with GPU acceleration
NVIDIA GPU-accelerated decision optimization has elevated the conversation around decision intelligence. How can GPUs improve decision optimization workflows? In what ways will solving problems such as workforce scheduling, vehicle routing, and energy optimization change? How does this change the way technology leaders think about their AI strategies?
NVIDIA cuOpt on Nextmv: Sync local runs, change compute, compare results, run scenario tests
See DecisionOps in practice with a vehicle routing problem (VRP) featuring NVIDIA cuOpt's routing engine on Nextmv using NVIDIA GPUs. This snippet comes from an NVIDIA + Nextmv panel discussion featuring a demo of cuOpt and Nextmv. In it, Ryan begins by syncing local cuOpt runs to a centralized UI in Nextmv for run tracking, reference, and collaboration.
Building stakeholder trust and confidence in decision intelligence
Human review and feedback is part of any decision workflow. From business leaders to front-line operators, stakeholder trust and buy-in can make or break the momentum behind any given decision intelligence project. So what can decision algorithm developers and teams do to increase the chances of success?
Lessons lived and learned: Project success and failure in decision science
While good stories of operations research and data science can come from practitioners of all kinds, this panel discussion encourages you to put your chin in your hands and elbows on the table and listen to the lessons lived and learned from a pair of industry leaders who have seen much and done much.
Open source in OR: Q&A with Pyomo and HiGHS
The operations research and decision science space has a diverse portfolio of open source projects, including Pyomo and HiGHS. Recently, new momentum is building around project adoption in research and industry contexts and ways to contribute and support to these communities.
The path to production: Exploring the software interface with OR
If you work on decision science projects, it is possible you’ve asked or been asked this question. Integrating a decision model into a live business process is a journey that necessitates applying software engineering practices to mathematical modeling in ways that are not always intuitive or straightforward. What are ways to demystify and streamline that process?
