Student Innovation & Entrepreneurship | Tampa, Florida
USF Student’s AI Startup Is Taking on One of America’s Most Expensive Problems: Cargo Theft and Freight Fraud
By [Staff Reporter] | Published: [2024] | The Oracle, University of South Florida
Freight fraud in the United States is not a minor inconvenience. According to industry data, fraudulent activity targeting carriers, shippers, and brokers jumped 27% year-over-year in 2024, with individual load losses frequently exceeding $40,000. For an industry that moves the backbone of the American economy, it is a crisis hiding in plain sight — and a USF student thinks he has the technology to stop it.
Abdulkhamid Abdullaev, an international student from Uzbekistan completing his undergraduate studies at USF and preparing to enter the university’s Master’s program in Entrepreneurship in Applied Technologies, is the founder of Dieselly.ai — an artificial intelligence company building what he calls a “fraud defense system” for the U.S. transportation industry.
The platform uses machine learning to continuously analyze data flowing from truck telematics systems, fuel cards, load boards, GPS trackers, and financial platforms, looking for the behavioral patterns that indicate double brokering, identity fraud, fuel theft, and manipulation of carrier contracts. Rather than waiting for losses to be reported, Dieselly.ai’s AI flags anomalies in real time — before money leaves the account and before cargo leaves the dock.
“Cargo theft and freight fraud are not random crimes,” Abdullaev explained. “They follow patterns. Bad actors repeat the same behaviors across different loads, different routes, different companies. AI can see those patterns. Humans reviewing individual transactions cannot.”
Abdullaev’s path to this startup is unconventional. He grew up in Uzbekistan, where his early career included serving as Tech Lead at the national Ministry of Higher Education, building data infrastructure for government institutions, and advising technology startups at Uzbekistan’s national IT accelerator. He was awarded a prestigious government scholarship to study in the United States, arriving with a clear thesis: that the same principles of data-driven anomaly detection he had applied to government systems could be applied to commercial transportation with potentially much larger impact.
His thesis proved out quickly. Abdullaev built and scaled a trucking company, Majha Transport LLC, to over seven million dollars in revenue during his Optional Practical Training period — and found, in the process, that every inefficiency and financial loss he encountered in his own operation was a product of the same structural problem: fragmented data that no one was connecting or watching.
Dieselly.ai now counts over 50,000 carriers scored on its Freight Intelligence platform and has flagged $8.1 million in freight fraud across its network. The company is integrated with major industry tools including Samsara, Motive, DAT, Truckstop, and QuickBooks, and is currently developing an AI-powered legal assistant that will help carriers and brokers generate and manage contracts electronically — closing another major vulnerability in the fraud chain.
For Abdullaev, the mission is personal as much as commercial. “Drivers lose their livelihoods to double brokering. Companies go bankrupt because of cargo theft. These are real people. If the technology exists to prevent it — and it does — then someone has to build it properly and put it in the hands of the people who need it.”
Abdullaev will begin his Master’s program in Entrepreneurship in Applied Technologies at USF this year, where he will continue developing Dieselly.ai alongside his studies. He credits the university’s entrepreneurship faculty with providing a framework that has helped him think about scaling an AI company as rigorously as he thinks about building one.
“USF has given me the language and the structure to think about growth at scale,” he said. “But the problem I’m solving — that I’ve always been solving — is the same one: how do you make institutions harder to defraud? The answer has always been data.”
Dieselly.ai | dieselly.ai | University of South Florida, Tampa, FL
