Large truckload carriers see driver turnover rates of 90 to 95% a year, meaning a fleet can expect to replace nearly its entire workforce every twelve months. That's the world Cargon AI is building for.
Khon Sultan (Tandon '26) is a former competitive wrestler from Uzbekistan who built a computer science degree, then a startup, out of an NYU application that arrived a day after the deadline. Azimkhon Sultankhodjaev, Khon's older brother and a Stanford grad, was already running his own startup when Khon asked to borrow a teammate, Edwin Liang. Liang left an engineering role at a West Coast startup to help build what came next.
It took a few wrong turns to find the right route. Cargon AI began as a grocery inventory tool, then a way to help brands recover money from retailers, before the team realized the real bottleneck was speed: trucking companies were losing drivers to whoever reached them first. Now, Cargon AI's AI agent reaches applicants within minutes, screens them, and gets them an offer before a competitor calls.
Tell us about yourselves, apart from Cargon AI.
Khon: Before startups, wrestling was a huge part of my life and taught me discipline, resilience, and how to keep going under pressure. In 2021, I left my family and moved to the U.S. to study computer science. It was exciting and difficult, but I embraced the challenge and finished college in two years. My journey from Uzbekistan to building a new life here shaped my ambition and who I am today.
Edwin: I studied computer science at the University of Michigan. A lot of my classmates went into big tech or hardware. I knew that wasn't the path I wanted. I also knew I couldn't build a startup by myself and didn't have an idea yet, so I did the unglamorous thing: pulled a list of companies YC and other accelerators had funded and cold-emailed my way onto a founding engineering team at one of them. My two co-founders there were an ex-Google engineer and a construction industry CEO. That job taught me the actual job of a founder isn't building, it's understanding what your customer needs.
Azimkhon:
I’m a graduate student in Management Science & Engineering at Stanford. I started as a machine learning engineer, working on forecasting models with EEG data, and later moved into the hospitality supply chain, where I worked on RFID systems to track inventory and improve visibility across manufacturers, distributors, and hotels. I like working at startups because I think they are one of the best places to learn, especially when the company is young and growing quickly. Working in startups taught me that almost every decision has to be filtered through growth. My time in hospitality supply chain also made me interested in how much operational friction comes from simply keeping different parties coordinated.
Who's Cargon AI for, and what does it actually do?
Edwin: Finding and hiring commercial truck drivers is expensive. We help trucking companies find new leads and hire them faster.
Khon: If a truck is sitting empty, that fleet is losing thousands of dollars a week. We built an AI agent that reaches out to driver applicants within minutes, screens them, and gets them hired before a competitor calls first.
How has Cargon changed from when you first started working together?
Khon: We started in retail, trying to help grocery stores manage inventory. Then we pivoted to helping consumer brands recover money that retailers owed them. We learned the hard way that we were in the wrong market. If your market can't pay, it doesn't matter what you build. Show me your market and I'll tell you who you're going to become. That's the truth we learned the hard way.
Was there a specific moment that changed how you think about the business?
Khon: When we picked up our SWOT analysis at the start of the program, the mentors basically told us: you guys know how to build, but you don't know how to make a business. Whenever you see a problem, you build something, but you don't think about whether it can actually scale. That one stung, but they were right. Nothing we'd built before was ever built at scale.
Edwin: There's a concept in machine learning called reinforcement learning, where a system makes mistakes over and over until it finds the version that actually works. That's a startup. You iterate through so many mistakes until you find the reward. The reward is the market, the right product, and the right customers.
You each described what AI means for this industry a little differently. What's the actual read?
Khon: Honestly, I think this industry needs labor support. The job is hard, drivers are tired, and fewer people want to do it.
Edwin: It’s transitioning this type of labor. People need to move from the physical labor part of the job to something that supports it instead.
What surprises you most about the industry you're working in?
Azim: Innovation. People think trucking is a traditional, old-school industry, but in the next two or three years I think it'll be dominated by AI: autonomous trucking, agents working across departments, a lot of the maintenance and hardware side too.
Khon: How hard the job actually is on people. If someone you love drives long-haul, you might not see them for four weeks at a time, and then they're only home for five days before they're out again.
Edwin: People don't give truck drivers enough credit. They're the reason your package shows up, and they're not seeing their families for a month to get it to you. We want technology to make that job easier, not just cheaper.
What's your biggest piece of advice for a student thinking about starting something?
Khon: Show me your market before you fall in love with your idea. We had four or five ideas before this one. The ones that failed weren't bad ideas, they just didn't have a market willing to pay for them.
What's next for Cargon AI?
Cargon is going to become a full stack ai native carrier helping brokers and manufactures to provide better service and infrastructure.
