Brian Sowul came into the trucking tech space by way of firefighting. He got his hazardous materials certificate on the job, and after discovering his fear of fire, he moved into the safety department at Swift Transportation, where he discovered his hatred of paperwork.
Now he’s senior product director at Transflo where he helped launch the company’s AI-powered workflow automation platform tailored exclusively for carriers. Sowul announced the new Workflow AI this week at the annual Truckload Carriers Association convention in Pheonix.
[RELATED: Pennsylvania logistics company launches digital freight marketplace tool]
“The paperwork for anything transportation is numbingly annoying – safety, revenue services, billing – you name it, there’s a document for it,” Sowul said. “We have come such a far way in the transportation industry, but in reality, there’s still so much paperwork because every time you have a line item, that line item correlates to an additional accessorial charge. That accessorial charge … require(s) a receipt … right now there’s men and women sitting in an office scanning documents or opening emails from those drivers.”
Transflo’s new tool utilizes AI to optimize that process so a human doesn’t have to review every single document.
“It’s going to process your documents at 2:00 in the morning or on a Sunday afternoon. Someone doesn’t have to be sitting at their desk to do it,” said Transflo Vice President of solution engineering Danny Goldstein.
Transflow started developing the tool and training AI with data from the millions of loads that pass through its system about three or four years ago, but it was set up for brokers. Processing payments to carriers was previously a very manual process for brokers when they would receive documentation.
“We’re going to do the same thing we did for brokers for carriers,” Goldstein said. “There’s enough information in the database and in the (transportation management systems) and rating engine stuff where we can make sure that the AI can make those decisions that don’t have to be tribal.”
The platform, designed to streamline document processing, reduces invoice lag and enhance operational efficiency, eliminates manual data entry, accelerates speed to cash and provides carriers with back-office automation. Workflow AI’s carrier beta testers have experienced a significant reduction in invoice lag, cutting turnaround time up to 80%, and a savings of more than 15 hours a week. The system has achieved up to 75% no-touch document readiness.
For every document a driver sends to the back office, the AI extracts billing-related data with multi-channel document ingestion and detects anomalies, resolving exceptions on its own or flagging it for human interaction.
“Instead of Bob coming in in the morning and capturing the document, indexing the document (and) auditing the document, it just goes into invoice status. At that invoice status, they just review it, hit approve, and it goes out,” Sowul said. “Imagine your processors, your billers, your AP clerks … they can now be redirected to something more meaningful and even drive higher revenue numbers somewhere else other than spending time looking at a document.”
The tool has validation methods in place, and if certain thresholds aren’t met or within an acceptable confidence meter, the AI flags it as an exception and kicks it over to a human for approval.
“Like, someone says Wal Mart, another person says Wal-Mart or Walmart together, or Walmart with 1234, or Walmart distro 12345. In reality, a human can quickly look at it and say, ‘Ah, that makes sense,’ but the AI says, I think it makes sense, but I would like you to look at it,” Sowul said.
Even though someone has to review it, he said the AI still reduces the time of that process from 20 minutes of reviewing a document to maybe a minute or two because the system flags the particular area that needs attention.
And Goldstein said it will only get better as Transflo ingests more data.
“When Brian first started working on Workflow, AI, the first day you plug it in, it works 25% of the time. Then six months later, it works 70% of the time, and now we’re at 80-plus percent of the time, and it’s working, so volume matters,” he said. “AI never replaces humans. We’re trying to be more efficient … sthose people that are doing 10 jobs can now do seven jobs. If you ever walked into a trucking office … we have some pretty overwhelmed people in the industry right now.”
Angel Coker Jones is a senior editor of Commercial Carrier Journal, covering the technology, safety and business segments. In her free time, she enjoys hiking and kayaking, horseback riding, foraging for medicinal plants and napping. She also enjoys traveling to new places to try local food, beer and wine. Reach her at [email protected].
