AI platform cuts delivery routing time from 20 minutes to two and a half
Logistics firm OneRail has launched OmniStar, built on Nvidia's AI software, to help retailers pick the cheapest and fastest delivery option for each order in near-real time.

Key points
- OneRail launched OmniStar, an AI-powered delivery routing platform built with Nvidia software, in 2025.
- The platform cuts the time needed to choose a delivery route from 20 minutes to roughly two and a half minutes, according to OneRail.
- OneRail's network covers more than 12 million drivers and over 1,000 logistics partners, and that data trains the AI.
- One early customer, a large tire distributor, is on track to save $40 million over three years by using the platform.
- OneRail projects OmniStar will handle more than $6 billion in gross merchandise volume (the total value of goods shipped through it) in the fourth quarter of 2025.
Small and mid-sized retailers have a delivery problem. Amazon and Walmart can route a package to your door cheaply and quickly because they have enormous teams, proprietary software, and years of shipping data. Everyone else has been relying on spreadsheets, phone calls, and guesswork.
OneRail, a logistics technology company based in the United States, says its new platform changes that.
What does OmniStar actually do?
OmniStar takes every available delivery option for a given order and scores them automatically, then picks the best one on cost, speed, and service level. It does this in about two and a half minutes. The same task done manually used to take around 20 minutes.
The platform runs on Nvidia's AI software and hardware, the specialised computing technology that powers many of today's large AI systems. OneRail told CNBC Tech it has been working with Nvidia for three years to build this capability.
"If you don't have the ability to make lightning-fast decisions, you're giving up margin," OneRail CEO Bill Catania said. "Last-mile fulfillment is expensive."
Last-mile fulfillment is the final leg of a delivery: the journey from a warehouse or sorting hub to a customer's front door. It is typically the costliest and most complex part of shipping.
Does it actually save money?
For one early customer, it appears so. A large tire distributor that piloted OmniStar is projected to save $40 million over three years by using its existing resources more efficiently. That is a company claim, not a peer-reviewed result, and the customer was not named.
OneRail also says the platform will exceed $6 billion in gross merchandise volume in the fourth quarter of 2025. That figure reflects how much product flows through the system, not OneRail's own revenue.
Azita Martin, Nvidia's vice president for retail and consumer goods, described the output as "a real-time decision layer that can route an order to the right carrier and delivery mode at the right cost, rather than relying on static rules or manual planning."
What does this mean for shoppers?
Directly, not much yet. OmniStar is a back-end tool for retailers and logistics managers. But if smaller retailers can genuinely cut delivery costs, the downstream effect could be faster shipping options and fewer "out for delivery" delays from businesses that previously could not afford tight logistics operations.
OneRail also has a partnership with FedEx, announced earlier this year, to offer same-day delivery across its customer base. Catania says OmniStar and the FedEx tie-up together give smaller businesses a realistic path to competing with the big-box giants on speed.
The claims here are still largely from OneRail itself. Independent audits of the platform's performance do not appear to be public yet. Watch for third-party case studies before treating the savings figures as settled.



