Information
-Sathish Raman
A significant share of the drivers keeping freight moving around the world do not speak the dominant business language used in the industry as a first language. For years, this has been an everyday problem in the freight industry around the world, but for a long time almost nobody chose to solve this language barrier issue. So drivers have done what they must: work harder to keep up, or rely on family members and dispatchers to translate for them.
Hey Bubba’s multilingual AI empowers truckers, dissolving language barriers. Drivers speak their native language to negotiate loads and manage dispatch hands-free. This innovative voice system creates an inclusive freight industry, unlocking talent and streamlining operations for a diverse global workforce.
Aswin Chandrasekaran wanted to build software that met those drivers where they were, in the language they actually speak.
As co-founder and chief technology officer of Hey Bubba, an AI dispatch company, he led the development of a voice system that lets a driver speak in a native language and have that speech translated, in real time and with low latency, into the industry-standard English a broker expects to hear during a rate negotiation. The company says it is among the first such pipelines built specifically for the noise and pace of live freight conversations.
“There are hundreds of thousands of drivers who are exceptional at the actual job and who were being held back by a language barrier that had nothing to do with driving,” Aswin said. “That never sat right with me. The technology exists to close that gap. Somebody just had to build it for this industry, in the conditions this industry actually works in.”
Those conditions are unforgiving for voice technology. A truck cab is loud, and cell coverage drops out for long stretches of interstate, so anything built for the road has to keep working with little or no connection. The harder problem, Aswin says, was the multilingual layer sitting on top of it: translating a driver’s words quickly enough to hold a natural back-and-forth with a broker who may have no idea the person on the line is speaking through an AI. Getting that translation to happen in the rhythm of a live negotiation, rather than the delayed cadence of a chat assistant, was the part no prior trucking software had attempted.
The result, is a dispatch assistant that a driver can talk to hands-free while on the road, one that can search for loads, negotiate, and handle invoicing without requiring the driver to type, tap through menus, or switch languages. For a driver who has spent years routing every phone call and rate discussion through someone else, the effect is less a new gadget than the removal of a daily obstacle.
For Aswin, the work carries a personal logic. Software that quietly assumes a single language, and a single way of working, tends to leave capable people on the outside, he says, a pattern he has watched repeat across every industry he has built in.
“I have spent my whole career around people who are brilliant but who do not fit the default assumptions a piece of software was built around,” he said. “When you widen those assumptions, you do not just help a few people at the margins. You unlock a huge amount of talent that was there the entire time.”
Aswin is quick to say the multilingual work is not finished. More languages still need to be added, and accents, regional dialects, and the many different ways people speak on the road remain a tough, ongoing engineering challenge. But he sees inclusion not as a nice extra, but as a test of whether the technology is any good in the first place.
“If your AI only works for the person it was most convenient to design for, it is not really intelligent yet,” he said. “The whole point is to meet people in their own language, in their own environment, and to do the work for them. That is the standard I want to be held to.”
