CJ Logistics runs agentic AI across 40+
warehouses.

The time to value that we've seen was almost instantaneous. We were able to build four dashboards in a matter of forty five minutes through conversational questions and answers. That was the power of the eye opening experience of how quickly data can be turned around into a dashboard, or actionable intel that drives value.

Then taking it to what it would look like if we had a brain that was out there thinking with us, looking over our shoulders and pointing us to opportunities.

It truly is a force multiplier when suddenly one person can do ten times as much as before and get all that visibility, be in the details, delegate the work, build systems that never forget, always optimize, and always strive for that next big win for you, even when you might not be looking or thinking about it. That's the exciting part.

Go to a big box retailer and just do a 360 after walking 30 seconds into the store, and ask yourself how everything arrived there in the right quantity, in the right condition, safely. Then walk 30 more feet and do it again. Old school was you sent people in to observe, sent an engineer in to observe. It doesn't matter what warehouse, even if it has automation in it, there are many lefts and rights you can make to do something in a different way. You can stage a pallet left or right, you can choose sometimes, or you are directed to store in other areas.

At the end of the day, operations used to be very reactive to a day of what's transpired, and a supervisor would have to spend a portion of their day digging through transaction logs to try and find the gap, and then ultimately hold a conversation on the next day or subsequent days with an operator. Now that's being delivered to them with video evidence to be able to show: here's the gap opportunity, here's your most help needed individuals, and here's some coaching feedback to make the most value back to the organization.

We can start to take some of the unviewed or unobserved tasks that historically we never would have even had a signal to go chase, and now go chase.

I saw what AI was doing in the world of software engineering, and then I saw our clients still living in spreadsheets and manual BI dashboards that had to be built over several weeks, months, sometimes years of projects to get reporting in place. And I was trying to figure out how we get these two worlds to work together, where we can give you the same kind of capability that we now have in software engineering, but in the world of operations.

I don't think there's been a QBR or a sales call over the last seven years that hasn't at least had some flavor of: here's one of our most important tools in safety. OneTrack has always been at the very top of that list, and now we're able to say OneTrack is at the top of that mountain for productivity, engineering and quality, on top of safety. CJ Logistics deepened our partnership with OneTrack because, speaking candidly, many technology solutions out there come with a system and an ability to put a lot of effort into it as a buyer of the software to get value. And in the OneTrack relationship, repeatedly, we have seen OneTrack put just as much effort into improving the outcomes as we put in ourselves. CJ Logistics views a partnership and an investment in a provider, and hopefully it's more partner than provider.

The greatest part about the service that OneTrack provides is that you are able and capable to do a lot of what we ask, and really help to guide us in the right direction. Sometimes we ask questions that you have to guide us back and say: well, what are you trying to solve?

Part of our belief in OneTrack is that you're able to wall off our information and keep it safe. Through our relationship, you've earned that trust. The AI is really something that we've been chasing, but allowing the OneTrack team to be an extra bandwidth of development and, quite frankly, thought leadership above and beyond our team, to start bringing expertise around what it would look like if we had a conversational chatbot that could give us an answer to a simple question.

Let's ask AiOn. Let's figure it out. And AiOn very quickly said: okay, I'm going to create a training page. Every time somebody signs into this page, they're going to go through the training of what this page is intended to solve and what this page is intended to show them.

I think that was one of the interesting learnings for us as we started to venture down this path of building AiOn and this agentic platform, because working with clients and close partners like CJ, we learned that it's about much more than just building a dashboard. You could go and use any kind of AI tool to put a dashboard together. But where it really becomes critical is if you're trying to create that secure layer that can sit on top of not just one, but multiple different LLMs.

It's compressed the time between an opportunity and a resolution. It's also enriched the level of data we have, and it really strengthened the platform to level up our conversations from what if, or we need to go investigate, to we know what's happening, and we can turn our people to value very fast.

And then you forced me out of my comfort zone and said: okay, you're going to type, you're going to tell it exactly what you want to build. I just dove headfirst into what was possible and very quickly became somewhat self sufficient. I spent a portion of my career in IT implementation, and this has probably been one of the easiest implementations because of the intuitiveness of the system and the software. In a few short prompts, a few short emails, we were able to not only get this cascaded out, but also measure the adoption to our entire network, and then roll that adoption up into group general manager and vice president level, and start to cascade some of that top down, bottom up approach.

Even starting from that point, now we have eyes on the floor, and from a vision perspective we can understand how a process is occurring. Then when you take that not just from a site, but to our entire network of facilities, thousands of employees, hundreds of forklifts, tens and twenties and thirties of buildings, you can start to not only solve a problem at a location that needs help specifically, but put together trends across a fleet of equipment and a group of awesome employees who all want to come in and do a great job that day, whether it's a supervisor on the floor or a wonderful frontline worker who's picking a case, or staging a pallet. And candidly, I think we've seen a correlation between sites that actively engage with the tool, ask good questions, not one prompt but three, four, five, six deep. The sites that are doing that are the top performing sites in our network. That's really allowed that flywheel of value to spin back towards the CJ team, and I think also the OneTrack team, to help move the product forward.

In a few quick clicks, I can get a network view, drop it down to a site view, drop it down to a shift view, drop it down to a supervisor view, and an individual view.

The more feedback we give it, the more robust the responses are, and I think that's the next phase for us: really training our entire workforce to embrace what AI is capable of. Every week I get a summary across our entire network of AiOn usage. It is probably one of the most exciting things that I get to read in my inbox, seeing which sites are leveraging the tool and what types of questions are being asked.

So in other words, the AI is helping itself get implemented across the network.

It's not all here to replace jobs. It's here to take us to the most valuable work, give some hours back to the business so we can keep a customer happy, or engage with our employees, which really, people and customers, is what our business is all about. And what excites me about that is every employee has the ability to be creative. Every employee has the ability to critically think. Oftentimes it was data that they were lacking, or the ability to even use the data that we had to get to an opportunity.

The point of the AI agent is not to eliminate anyone's job. It's truly to give your team, the great operators that you have in the field, the very best tools so they can be as effective as possible every single day. Because ultimately they don't have to spend the time in front of the computer doing the computer work. They can spend the time with the people, doing the really important people work in the field with their team, developing their employees.

If you ask anyone who's run a warehouse, people are at the core of it. When we say managing, it's really helping people get through their day easier, removing obstacles. That's what keeps people coming back to an operation. They walk up to the door and they say: I like working here because the team around me helps me solve problems, I'm not having to do it by myself every day.

We really changed from a reactive post shift investigation, or post shift opportunity, to autonomous real time floor enforcement. Our supervisors are no longer spending hours digging into data to get the decisions. It's being driven directly to them so that they can spend the most time where it counts, with the operators out on the floor driving value. From an annualized perspective we are tracking on target right now. Our clock in, clock out gap opportunity has dropped by forty five percent. From a productivity perspective, our units per hour performance has increased by eighteen percent, in comparison to really the scope and scale of headcount. When I look at a comparative rollout of a BI platform, or a data and analytics platform previously, we easily had three to five times the support staff specifically on the project.

As it relates to what it's allowed us to do in our business, I do think it has allowed us to think about other verticals. We were historically maybe focused in two or three verticals, consumer goods, electronics. Now the power of having these tools is allowing us to learn more about our operations day to day, and start to have more confidence in taking on a new vertical, and how quickly we could learn about the productivity and the quality as it relates to that. And I think with the power of these tools, we're starting to have more confidence that we can get that into our muscle memory at ten times what it would have been in the past, and that might be an understatement. As we stack ourselves up against our competition, we do feel like this path, this choice we've made to develop this way, puts us in a very strong position to be able to say that we have visibility across every operation, from the sensors, information feeding off of the platform that drives fast decision making, and the ability to drive that into value for a customer, whether that's service or cost. Back to creating value, we can say this is no longer a person dependent process. It's a system driven process that also has the ability to improve itself over time, and I don't know that all of our competitors are able to say that in the same way. Trying by yourself can be a little intimidating. Trying with a trusted partner can be very reassuring. I don't think we would have been here without OneTrack.

Agentic AI is no longer just a pilot or a future concept. It's living within our operations today, and we're utilizing that to drive real time automated enforcement and reporting as the new baseline for logistics excellence.