Lightnews — Scholar-powered news

In conversation about AI/ML + mathematical optimization with Yash Puranik, Director of Decision Sciences at Aimpoint Digital

1 2

Nextmv @nextmv.bsky.social · Sep 2

“Once you start learning how to recognize mathematical optimization use cases, they exist pretty much everywhere: in finance, logistics, supply chain, manufacturing, energy, and even government — the possibilities are endless.”
– Yash Puranik

Read the full interview >> hubs.la/Q03Fw-lR0

Explore the applications of mathematical optimization alongside the AI/ML landscape for predictive analytics, LLMs, and beyond.

Tracking decision model runs: an OR-Tools region allocation example in Python

Nextmv @nextmv.bsky.social · Aug 22

Check out this tutorial where we use OR-Tools with Nextmv to record allocation model runs, results, input data, and more: hubs.la/Q03DRHhC0

#decisionscience #datascience #orms #decisionops #decisionintelligence #ortools

Record your assignment model runs, results, input data, and more in Nextmv. Add a few lines of code to your existing model to track every run and easily share data with your teammates.

In conversation about ML + OR with Jerry Yurchisin, Senior Data Strategist at Gurobi

Nextmv @nextmv.bsky.social · Aug 12

What makes now a great time for combining #machinelearning and #operationsresearch?

Ryan O’Neil, CTO at Nextmv, recently sat down with Jerry Yurchisin, Senior Data Science Strategist at Gurobi, to discuss the opportunities for combining them, challenges in adoption, and more. hubs.la/Q03Cm9BV0

This interview explores perspectives on the intersection of ML and OR, challenges and opportunities for data scientists in the optimization space, observations in industry, and why now is a great time...

4 uses for optimization in software

Nextmv @nextmv.bsky.social · Jul 17

Optimization is often highlighted in routing, scheduling, inventory, and other tangible, visible contexts. In this post, @ryanjoneil.bsky.social explores use cases in computing and software: hubs.la/Q03xJjF40

#orms #decisionscience #decisionOps #sql #informs #optimization #operationsresearch

Optimization is often highlighted in supply chain contexts – routing, scheduling, inventory, and other tangible, visible problems. This post explores a few use cases in computing and software.

How to record decision model results, visuals, and I/O with tracked runs: A facility location example

Nextmv @nextmv.bsky.social · Jul 16

Decision model development and collaboration is easier when there’s a shared system of record to reference and interact with for I/O, results, and charts. Learn how to create one in Nextmv using tracked runs: hubs.la/Q03xtJdw0

#orms #decisionOps #decisionintelligence #operationsresearch

Decision model development and collaboration is easier and faster when there’s a shared system of record to reference and interact with for I/O, results, and charts. Learn how to create one in Nextmv.

Nextmv FICO® Xpress Integration: Accelerating optimization teams with modern DecisionOps

Nextmv @nextmv.bsky.social · Jul 8

👋 Say hello to the Nextmv FICO Xpress Integration! See how DecisionOps accelerates optimization teams with decision model development, deployment, testing, monitoring, and collaboration. Read on for more details: hubs.la/Q03wh09Z0

#orms #decisionOps #decisionintelligence #operationsresearch

We’ve partnered with FICO® Xpress Solver to provide a modern DecisionOps platform experience to FICO Xpress users, allowing them to create end-to-end decision pipelines for model deployment, testing, ...

Quickly compare runs of your decision model in the Nextmv UI

Nextmv @nextmv.bsky.social · Jun 26

“Did the latest run of our routing model utilize fewer vehicles than the run we made yesterday?”

“Did we configure the same stop balancing constraint for those two runs?”

Answer these questions in a few clicks by comparing decision model runs in the Nextmv UI. hubs.la/Q03tPH1L0

#decisionscience

Select multiple runs from your model’s run history for a side-by-side comparison of their KPIs and model configuration. Answer performance questions in a few clicks and determine what to test next to ...

In conversation about DecisionOps with Fred Gardi, founder and CEO of Hexaly

Nextmv @nextmv.bsky.social · Jun 10

Revisit a conversation with Fred Gardi and @ryanjoneil.bsky.social covering perspectives on growth/adoption of decision science, advocating for optimization within broader AI strategies, and the Nextmv Hexaly integration. 👇👇👇

#decisionscience #DecisionOps #hexaly #datascience #heuristics #orms

This interview explores perspectives on growth and adoption of optimization across practitioners, challenges and opportunities in the optimization space, thinking about optimization in broader AI stra...

Simplify optimization model run queueing with Nextmv's DecisionOps platform

Nextmv @nextmv.bsky.social · May 22

Create and manage queues directly in Nextmv to easily prioritize optimization model runs. Configure prioritization at runtime to prevent blocking the queue with large runs or at the model version level to ensure production runs have precedence.

www.nextmv.io/blog/simplif...

#decisionscience

Create and manage queues directly in Nextmv to easily prioritize model runs. Configure prioritization at runtime to prevent blocking the queue with large runs or at the model version level to ensure p...

How Veho switched to a new optimization solver in under one month with DecisionOps

Nextmv @nextmv.bsky.social · May 5

Learn how Veho launched a new optimization solver in under a month with Nextmv.

“The beauty of Nextmv was that we didn’t have to change anything in our system to call the new solver — it was effectively plug-and-play... We basically needed zero engineering help.”

When Veho’s data science team wanted to evolve their route optimization with a new solver, they leveraged Nextmv’s DecisionOps platform to transition to Hexaly in a matter of weeks.

Automatically expose native Gurobipy and scikit-learn parameters as options in Nextmv

Nextmv @nextmv.bsky.social · May 1

Tinkering with the Gurobi Optimization “Heuristics” option? Or the scikit-learn “Strategy” option? Two new packages from Nextmv automatically expose native parameters as configurable options so you can tune your models in a few clicks. Learn more >> www.nextmv.io/blog/automat...

#decisionscience

Tune your Gurobipy and scikit-learn models with configurable options so your team can prototype faster, experiment with ease, and manage their models with more insight on Nextmv’s DecisionOps platform...

Build decision workflows to automate multiple steps in an optimization pipeline

Nextmv @nextmv.bsky.social · Apr 23

Looking for an MLflow experience for optimization? Create end-to-end decision workflows in Nextmv with steps such as data prep, regression, optimization, and visualization. Learn more >> www.nextmv.io/blog/build-d...

#decisionops #mlops #logistics #supplychain #decisionscience #datascience #orms

Looking for an MLflow experience for optimization? Create and manage end-to-end decision workflows in Nextmv with steps such as data prep, regression, optimization, and visualization. Press play to ki...

Replay optimization model runs under different conditions with scenario testing

Nextmv @nextmv.bsky.social · Apr 17

Build and run an ML + OR decision workflow in Nextmv: www.nextmv.io/blog/how-to-...

Prepare data (Pandas) → run regressor (Statsmodels) → run optimizer (Gurobi) → visualize output (Plotly) and then run a scenario test!

#orms #DecisionOps #DecisionFlow #python #datascience #MLFlow #MLOps #informs

Nextmv @nextmv.bsky.social · Apr 3

🎛️ Troubleshooting an operational issue or a poorly performing plan? Identify better plans by replaying your run with varying settings. Create multiple scenarios to understand the impact of input data, model configuration, and more on your KPIs in Nextmv’s UI. nextmv.webflow.io/blog/replay-...

Identify better plans by replaying your run with varying settings. Create multiple scenarios to understand the impact of input data, model configuration, and more on your KPIs in Nextmv’s UI.

nextmv.webflow.io

Nextmv @nextmv.bsky.social · Mar 26

Optimization is key to MatchBack Systems minimizing costs and improving equipment utilization for its customers — but it's the speed to deployment and model iteration that takes their solution to the next level: hubs.la/Q03dtMKg0

#orms #supplychain #datascience #decisionscience #decisionOps

MatchBack Systems: Eliminating historical drayage inefficiencies for imports and exports

Optimization plays a key role in MatchBack Systems successfully minimizing costs and improving equipment utilization for its customers — but it's the speed to deployment and model iteration that takes...

Render custom visualizations for your optimization model with Plotly, Chart.js, and GeoJSON

Nextmv @nextmv.bsky.social · Mar 6

🎨 Using Plotly, Chart.js, or GeoJSON to visualize data from your optimization models?

Streamline analysis and collaboration with interactive charts, plots, and maps that automatically render in the context of run details. www.nextmv.io/blog/render-...

#datascience #orms #logistics #supplychain

Standardize and manage custom visuals as data assets in Nextmv alongside your model. Accelerate your workflows with interactive charts, plots, and maps that automatically render alongside run details ...

Nextmv Hexaly Integration: How to run, test, and manage with DecisionOps workflows

1 1

Nextmv @nextmv.bsky.social · Mar 3

Interested in Hexaly to solve optimization problems? Thursday's techtalk will show how to get started with Hexaly on Nextmv, use features for scenario testing and best plan auto-selection, plus Q&A with Hexaly’s CEO: www.nextmv.io/videos/nextm...

Interested in Hexaly to solve optimization problems? This techtalk covers how to get started with Hexaly in Nextmv, use features for scenario testing, best plan auto-selection, model management, and a...

Reproduce and replay optimization model runs to troubleshoot operational issues faster

Nextmv @nextmv.bsky.social · Feb 25

“I’m putting the schedule together for tomorrow. Why are we using fewer drivers?”

When an operational concern is raised, reproducing it is critical. With Nextmv, replaying optimization model history (with easy access to all the data) is just a click away.

www.nextmv.io/blog/reprodu...

When an operational issue is reported, reproducing it is one of the first steps in an investigation. Finding and connecting the data you need to triage the issue can be intensive. With Nextmv, replayi...

Deploy Python decision models as hosted apps directly from a Jupyter Notebook

Nextmv @nextmv.bsky.social · Jan 29

Building a decision model in a Jupyter Notebook? Push your Python model from a local file to a remote app with Nextmv. Conduct tests, collaborate with teammates, and get observability into model performance.

www.nextmv.io/blog/deploy-...

#decisionscience #datascience #orms #python #jupyter

Push your Python decision model from a local file to a remote application in minutes – whether you’re using a notebook or running in another Python environment. Conduct tests with fully featured exper...

Auto-select the best plan from your optimization when comparing decision models and solvers

Nextmv @nextmv.bsky.social · Jan 23

With new ensemble runs on Nextmv, you can evaluate multiple models and solvers to automatically surface the best performing plan – with all the supporting context you need to share results. www.nextmv.io/blog/auto-se...

#decisionscience #datascience #orms #optimization #simulation

Run multiple solvers or models in parallel to automatically select the best plan for your business with ensemble runs. Use your unique rules as selection criteria to easily converge on a plan, validat...

A 2024 lookback and 2025 preview of DecisionOps workflows

Nextmv @nextmv.bsky.social · Jan 15

👟 We’re hitting the ground running in 2025! We look back at how the Nextmv DecisionOps platform evolved over the last year and what we plan to release this year. www.nextmv.io/blog/a-2024-...

#decisionops #machinelearning #datascience #decisionscience #orms #python

2024 was the year of the Python modeling experience, optimization integrations, and helping modelers find solutions even faster with parallel runs and interactive visualizations. In 2025, we look forw...

Nextmv @nextmv.bsky.social · Dec 5

What can DS/OR practitioners do to efficiently deliver more decision model code and model value for real-world impact? A first step, adopt modern Git practices: pubsonline.informs.org/do/10.1287/o...

#orms #informs #operationsresearch #decisionscience #analytics #cicd

Optimization hackathons, code fests, and innovation challenges on Nextmv

Nextmv @nextmv.bsky.social · Nov 19

Looking to innovate in operations research? Or better translate academic research to industrial practice? Build modeling skills set up for real-world impact? Check out this starter guide to hosting challenges on Nextmv. www.nextmv.io/blog/optimiz...

Looking to innovate in operations research? Or better translate academic research to industrial practice? Build or practice modeling skills set up for real-world impact? Check out this starter guide t...

Nextmv @nextmv.bsky.social · Nov 15

👋