OneRail Launches AI Delivery Platform Powered by Nvidia Technology

 

AI is now helping companies decide not just where a package should go, but how it should get there.

OneRail has launched OmniSTAR, an AI-powered delivery optimization platform that uses Nvidia technology to help retailers, wholesalers, and distributors determine the best way to fulfill individual orders.

Instead of relying on a fixed delivery method, OmniSTAR can compare several options, including a company’s own fleet, local couriers, parcel carriers, and other transportation modes. The system then selects the lowest-cost option that can still meet the required delivery service level.

At the core of the platform are Nvidia’s cuOpt optimization engine and cuDF GPU-accelerated data processing technology, combined with OneRail’s own pricing, delivery performance, and logistics data.

According to OneRail, the technology can dramatically shorten the time needed to calculate delivery scenarios.

A process that previously took around twenty minutes can now reportedly be completed in less than two minutes. More complex calculations that could take as long as one week may be reduced to approximately two days.

That speed matters because delivery decisions often need to happen while orders are already moving through live fulfillment operations.

OneRail CEO Bill Catania told CNBC that companies risk losing profit margins when they cannot make these decisions quickly, especially because last-mile delivery remains one of the most expensive parts of logistics.

AI That Predicts, Then Decides

OneRail uses different forms of AI throughout the delivery process.

Its machine-learning models can estimate variables such as expected delivery time, the probability of delays, the likelihood of a successful first delivery attempt, and possible pricing ranges.

Those predictions can then feed into optimization systems that decide how a specific order should be handled.

OmniSTAR goes a step further by comparing different fulfillment modes and choosing the option that offers the best balance between cost and service requirements.

This reflects a broader trend in logistics AI.

Prediction tells a company what is likely to happen.

Optimization helps decide what the company should do next.

That distinction is important because real-world logistics constantly changes. Traffic, weather, fuel prices, driver availability, road closures, equipment problems, and new urgent orders can all affect the best delivery decision.

Nvidia cuOpt Powers the Optimization

Nvidia describes cuOpt as a GPU-accelerated optimization library designed for problems such as vehicle routing and mathematical optimization.

The technology can account for multiple variables, including vehicle capacity, travel time, operating schedules, starting locations, transportation costs, and other logistical restrictions.

Instead of checking every possible delivery combination, which could take enormous computing resources, cuOpt generates potential solutions and continuously improves them using GPU-accelerated optimization techniques.

For OmniSTAR, OneRail said cuOpt is being used not only for route planning but also for selecting the most appropriate delivery mode.

The platform also uses Nvidia cuDF, which accelerates the processing of large datasets such as pricing records, delivery histories, and operational performance data.

OneRail combines Nvidia's technology with its own logistics data gathered from millions of deliveries.

The company said its network includes more than twelve million drivers and over one thousand logistics partners.

Using that information, OmniSTAR can analyze how different delivery decisions affect cost, service performance, and even the profitability of individual products.

AI Can Recalculate When Conditions Change

One of the most important capabilities of the system is reoptimization.

A delivery plan that looks efficient in the morning may no longer be the best option several hours later.

A vehicle could break down.

Traffic conditions could suddenly worsen.

Fuel prices could change.

Bad weather could disrupt routes.

A high-priority order could unexpectedly enter the system.

Because Nvidia cuOpt is stateless, updated conditions can be submitted as a new optimization problem so the system can calculate another solution.

OneRail said OmniSTAR can repeatedly evaluate delivery scenarios as business and operating conditions change.

Already Being Used by Major Companies

OmniSTAR is already being deployed by selected enterprise customers.

One example cited by OneRail is US Foods.

According to the company, OmniSTAR identified delivery arrangements that were hurting margins, including situations where relatively low-margin products were being transported over long distances using more expensive delivery equipment.

OneRail said US Foods used those findings to adjust pricing and restructure some of its delivery operations.

The company also told CNBC that an unnamed large tire distributor using the platform achieved approximately forty million dollars in run-rate savings over three years.

Because the customer was not publicly identified, that figure remains a claim provided by OneRail.

The company also expects OmniSTAR to process more than six billion dollars in gross merchandise value during the fourth quarter of twenty twenty-six.

According to CNBC, OneRail and Nvidia worked on the technology for approximately three years before its launch.

The collaboration reportedly included direct work with Nvidia’s cuOpt engineering team on large-scale logistics optimization and last-mile delivery.

Earlier this year, FedEx also launched FedEx SameDay Local in collaboration with OneRail, connecting customers with a nationwide network of more than one thousand delivery providers.

What This Means for the Future of Logistics

The bigger story here is not simply that AI can calculate faster routes.

AI is increasingly being used to make operational business decisions in real time.

For logistics companies, the question is shifting from:

“What is the fastest route?”

to:

“What is the smartest and most profitable way to deliver this specific order right now?”

That could mean choosing a company-owned truck for one delivery, a local courier for another, and a parcel carrier for the next, all decided automatically based on changing costs and service requirements.

For businesses operating on thin margins, even small improvements across millions of deliveries can translate into significant savings.

And this is another example of how AI is moving beyond chatbots and content generation.

It is increasingly becoming part of the invisible infrastructure making decisions behind transportation, supply chains, manufacturing, retail, and everyday commerce.

For the Philippines, this is worth watching closely.

As e-commerce, food delivery, retail logistics, and last-mile services continue to grow, AI-powered optimization could eventually help Filipino companies reduce delivery costs, improve route efficiency, and make better use of limited transportation resources.

The future of AI may not always look like a robot.

Sometimes, it could simply be the intelligence deciding which truck, rider, courier, or route should deliver your package.