Gemini is a huge game changer in terms of time. These tools are making processing of information that much faster.
Aayush Thakur
Co-founder
As a graduate student, Aayush Thakur was analyzing billions of trucking industry data points when he noticed puzzling clusters of trucks idling for up to 24 hours. He learned that truckers often get stuck due to driving time regulations, or they have to wait for a load to avoid returning with an empty trailer. This not only delays essential cargo and wastes fuel, it keeps drivers on the road longer. “Everything we use touches trucks, and the people doing all this hard work are spending time away from their loved ones,” Aayush notes. Determined to fix this, he patented his algorithms and with his wife and COO, Deme Yuan, founded FR8relay. The company’s software facilitates a relay model for brokers, so they can divide long routes into shorter segments, allowing drivers to swap trailers and return home daily. This reduces downtime by 20–30 percent. Today, the AI-powered platform optimizes nearly 20,000 loads annually through a partnership with a logistics broker.
FR8relay hosts their entire system on Google Cloud, turning raw data like shipping bids and locations into actionable logistics. Their proprietary Vertex AI agents analyze the data; instead of manual entry or running reports, users can simply ask a chat interface, “What was the rate on this route three months ago?” to get instant context and compare options in a single dashboard. And truckers can upload photos and documents directly to Cloud storage to speed up invoicing. Gemini parses brokers’ 3,000 daily emails, instantly summarizing complex rate quotes into clear bullet points and drafting replies that help keep deals moving. The company is testing a machine-learning tool that predicts delays by analyzing traffic and weather patterns. “Our goal is to make driving a day job,” Aayush says, “getting cargo where it needs to go faster and giving American truckers more time with their families.”
Economic Impact --- Arkansas