Like any new technology, agentic artificial intelligence has a dizzying array of potential applications. My job is to make sure that out of everything we could do with AI, we’re doing what we should do with AI.
As I said onstage at the NextGen Supply Chain Conference, I have a great responsibility as Chief Technology Officer to ensure C.H. Robinson’s 83,000 customers are getting the highest and most benefit from AI. Lean AI is the way we’re accomplishing that.
Our Lean operating model is how we run our entire business. Well-defined strategies, workstreams, and measurement create a cycle of continuous improvement. We’ve applied that same rigor to AI since we built our first AI agents in 2023. We’re using agentic AI where it can be scaled and have the most impact for customers, carriers, and employees alike. We’re disciplined in prioritizing which AI agents to build. Business results must be tangible and the ROI clear.
Disciplined choices lead to powerful outcomes
Our quoting agent is a prime example of our disciplined approach to AI. When a new form of AI came along and it was capable of performing tasks and making decisions like a human would, we could’ve tried to apply it to any number of complex business processes—or just built a cool logistics chatbot that was fun to talk with. Instead, we started by tackling the more than 10,000 emails we get every day requesting a price quote.
Not super flashy. But for every one of those 10,000 emails, a human had to open it, read it, figure out what the customer wanted, go to our Dynamic Pricing Engine for a quote, and send it to the customer. So we built an AI agent that can do all that in 32 seconds.
It’s operating at scale. Our agent has already delivered more than 1.5 million price quotes, and we’ve trained it to handle not only truckload price quotes, but also less-than-truckload, flatbed, and even hotshot loads.
It has tangible business impact. We’re serving our customers hours faster. The valuable time our logistics experts would’ve spent on 1.5 million price quotes is spent doing more strategic work for our customers. We’re also more competitive in the marketplace, because we’re bidding on—and winning—more freight.
Multiply that impact across all the tasks that our fleet of more than 30 AI agents is now doing: everything from orders to appointments to tracking to invoices and many other essential, formerly time-consuming tasks in the shipping process.
A future where AI acts more predictively and proactively
As exciting as our progress has been so far, I’m even more excited about what’s to come. The future I see for logistics is one where our world-class logisticians, supercharged by our world-class Lean AI, are operating even smarter, faster, better supply chains for our customers.
Our people are already learning to work with and oversee AI agents. They have more time to do what they do best: tackle the most complex logistics challenges our customers and world events throw at them. Our AI agents are already performing increasingly complex tasks. They’re already capable of more sophisticated reasoning and can solve more problems by orchestrating other AI agents.
Now, the same discipline and rigor we’ve used to create task-specific agents is being applied to the larger AI ecosystem we’re building so customers of any size or industry can have an Agentic Supply Chain – one that continuously thinks, learns, adapts, and acts. It will understand context and make decisions in real time. It will act more predictively and proactively.
As my team of more than 450 software engineers, machine learning engineers, and data scientists continues to push the envelope, they will be guided by Lean AI. Over a dozen investment banks have recognized us for our achievements with AI, and one of my favorites was the one that called us “AI substance in a landscape littered with façades.” That meant a lot to me, because it validated our whole approach to AI. It’s substance that put us ahead and will keep us ahead in AI.
