Startups & Entrepreneurs

Orlando Startup OneRail Launches OmniSTAR, an AI Delivery Platform Built With NVIDIA

Behind every online order lies a fleeting calculation over trucks, couriers and margins — a problem the Orlando logistics firm OneRail says its OmniSTAR platform, unveiled in September 2026 with NVIDIA and deployed with select enterprise customers, can compress from twenty minutes to under two.

4 minute read

ShareFacebookX

Behind every online order sits someone facing a quiet race against the clock: minutes to decide which truck, which courier and which delivery promise to make a customer, and at what cost. Make that call slowly, or out of habit, and the margin on the sale quietly vanishes. In Orlando, a company that has spent years assembling the data behind that decision has now put NVIDIA’s computing behind it — unveiling a platform it says can weigh every delivery option for an order and answer in under two minutes a question that used to take twenty.

The company, OneRail, introduced the platform — OmniSTAR — on Sept. 1, 2026, in a release via Business Wire. Developed with NVIDIA accelerated computing and acceleration software, the AI-powered delivery decisioning system evaluates every available option for each order — across owned fleet, courier, parcel and other modes — and selects the one that meets service requirements at the lowest cost, in real time. According to the company, it is already deployed with select enterprise customers.

The problem is arithmetic, not trucks

For retailers making thousands of delivery decisions a day, the bottleneck has been computational. Because conventional systems cannot evaluate the combinations quickly enough, most retailers still rely on static rules or manual planning, OneRail said in its release. OmniSTAR pairs NVIDIA’s cuOpt GPU-accelerated decision optimization engine and cuDF data processing with OneRail’s own pricing and performance dataset — compiled, according to the company, from millions of deliveries across a network of more than 12 million drivers and over 1,000 logistics partners. That multi-year collaboration, run through the NVIDIA Inception program with direct engagement from the cuOpt engineering team, cuts computation time by as much as ten times, in the company’s figures, compressing a 20-minute problem to under two minutes and a week-long calculation to roughly two days.

For Bill Catania, OneRail’s co-founder and CEO, the stakes come down to who can afford speed. “For most retailers, optimizing last-mile delivery at scale has not been feasible in real time. OmniSTAR changes that,” Mr. Catania said in the release. “Until now, same-day delivery at scale was a capability only large retailers like Amazon could afford to build.” Azita Martin, NVIDIA’s vice president and general manager for retail and consumer packaged goods, said: “OneRail’s AI-powered delivery platform demonstrates how AI is benefiting businesses and consumers. With CuOpt decision optimization engine and NVIDIA AI, OneRail addresses highly complex logistics challenges at scale, enabling faster delivery and lower logistics cost for retailers.”

Deployed, with one named result

So far, the release offers one detailed look at the model in practice. At US Foods, OneRail said, the system identified delivery configurations that were eroding margins — including low-margin items shipped long distances on high-cost equipment — insights the customer used to adjust pricing and restructure delivery patterns. Beyond routing orders, the platform’s stated role is diagnostic as much as logistical: flagging internal rules and constraints that drive unnecessary cost, showing how delivery decisions affect item-level profitability, and letting teams re-run strategies as fuel costs, weather and shipping conditions shift. The release publishes neither pricing nor a full customer roster, and this article asserts neither.

Orlando’s logistics layer, at platform scale

Headquartered in Orlando, Florida, with global operations, OneRail describes itself as the AI technology platform orchestrating unified commerce for enterprise retailers, wholesalers and distributors, backed by its OmniPoint platform and a 24/7 U.S.-based Exceptions Assist team. Its largest credential predates OmniSTAR: in March 2026, FedEx selected OneRail as the technology and network partner powering FedEx SameDay Local. That places the company alongside the region’s other platform-scale firms — the payments processor profiled in this desk’s Stax explainer and the Stax founders’ second company, Worth — businesses whose Orlando headcount matters less than the national infrastructure quietly directed from here. The broader landscape appears in this desk’s guide to starting up in Orlando.

Whether OmniSTAR reshapes last-mile economics more broadly now turns on adoption. A decisioning platform is ultimately worth what retailers change because of it, and OneRail has put just one named result on the record so far. The disclosures that would measure its reach — a fuller customer roster, published pricing, a general-availability date — were left out of the release, and this desk will report them when the company puts them on the record.

More from this desk: the Startups & Entrepreneurs hub and the full Startups archive.

Sources & notes

OneRail, “OneRail Launches OmniSTAR, the First AI-Powered Delivery Decisioning Platform Built with NVIDIA AI,” company announcement via Business Wire, Sept. 1, 2026 (businesswire.com), read in full Nov. 17, 2026 — the platform’s function and mode coverage, the NVIDIA cuOpt and cuDF components, the Inception collaboration, the company’s speed figures (as much as ten times faster; a 20-minute problem to under two minutes; a week-long calculation to roughly two days), the dataset figures (millions of deliveries; more than 12 million drivers; over 1,000 logistics partners), the select-customer deployment status, the US Foods example, the FedEx SameDay Local selection in March 2026, the OmniPoint platform and Exceptions Assist team, the Orlando headquarters, and both quotations, reproduced exactly as published. All performance and network figures are the company’s own and are attributed as such in the story. No pricing, funding, valuation or headcount is asserted, because the release publishes none. Featured image: Orlando Citizen media library (pipeline donor reuse of the site’s own published photography, recorded in the gate record).

Filed under · · · ·

Editor’s Picks

  1. 1NewsEpic Universe Faces Its First Holiday Season
  2. 2ReviewsThe 10 Best Coffee Shops in Orlando
  3. 3TourismMagic Kingdom
  4. 4Top BusinessesHaan Coffee: Mills 50’s Korean-Inflected Roaster Tops Orlando Coffee
  5. 5ReviewsThe 10 Best Attractions & Things To Do in Orlando

Spotted an error? Tell us — corrections · How we rank

Elena Marsh

Orlando Citizen startups and entrepreneurs reporter. Covers Central Florida founders, funding rounds, accelerators and the companies scaling out of Orlando, Lake Nona and NeoCity — reported from company announcements, investor and university releases and the public record, with every figure traced to its source.