EasyPost Insider: Luma AI Advisor, the Shipping LLM You Can Rely on With Tyler Diestel From EasyPost - Unboxing Logistics Ep. 85

Welcome to Unboxing Logistics. I'm your host, Lori Boyer of EasyPost. Today we've got a fan favorite. Back again, the one and only Tyler Diestel.

He's here for one of our EasyPost Insider edition episodes. That's where you get to learn all about what's happening in the product world at EasyPost and maybe learn a little bit something new about something coming out. And so that is what we're gonna be discussing today, and it is, you guys, one of my favorite releases we've ever had here at EasyPost. Totally geeking out over this one.

Tyler, introduce yourself. I'm sure everyone knows you, but for the random person who hasn't ever seen a Tyler Diestel episode, tell our our family here who you are. Hi, I'm Tyler Diestel. I'm a product manager of the data platform here at EasyPost.

I've been here for over six years and have throughout the entire time been focused on our data products. You know, you said one of my favorite words, data. So we are gonna be talking today about a, a new product, a new feature that's recently come out in conjunction with our data. It's Luma AI Advisor, and it is one of the first, one of the best LLMs designed specifically for the shipping industry.

Okay. I love LLMs because they make my life so much easier. I can just ask a question and be like, what am I supposed to do? And LLM is giving the answer.

So Tyler, gimme a little bit of background about maybe how the idea for Luma AI Advisor came about. We have this amazing product suite called Luma AI. It does a lot of things, but one of the biggest things that does is it gives our shippers insights into a treasure trove of data. The problem is a lot of times we provide too much data and it somewhat overwhelms our shippers, and they don't know what to do with it.

So with Luma advisor, users can now, instead of having to sift through the various dashboards that we provide, they can just chat with this nice little chat bot to find their answers directly. Yes. I love it. So for years with data, I feel like we've always been insights, insights, insights.

You can get new information, new data, but this helps you, what I'm hearing, make you give you decisions. Is that what you would say? Absolutely. This helps you reach the decisions you need to make.

It's not generic so that it answers everyone's basic question. It answers your specific question that you are looking for, and it's really tailored to the goals that you tell it. Yeah So why would a shipping specific LLM, so I, I, I guess I should back up. An LLM is a large language model.

We're talking Claude We're talking ChatGPT, we're talking Gemini, those kind of things, right? Why would somebody want. An LLM that is focused specifically for shipping versus just using some of the more generic LLMs. I would say using the LLM that's specific to your shipping habits is exactly what you wanna use over a generic ChatGPT or Claude.

Those are great, but those are answering high level, not really personable questions. This LOM has all your shipping data so it knows exactly what you have done and what you can do to get better, to reach your goals. Hmm. So it has your own data.

I, I've interviewed so many AI experts and they always say AI is only as good as the data it's trained on. 1000%. What data is this trained on then? Can you share some of that?

You said it was their data. This LLM is trained on your data, but It also takes advantage of our AI models that produce our time in transit predictions. That comes into play when we provide recommendations to you on, let's say, which carrier to switch to, to maintain the same time in transit criteria that you are used to or want to achieve. So how is it different than from just the traditional like rate shopping kind of tools we've had for a long time?

It's different in the sense that it has more information to provide you to get to your exact label that you're looking for. For example, if you're doing traditional rate shopping, there's only so many pieces of information you have at your disposal to say, hey, let me just purchase this next day air, because in and of itself, it says it will get there in the next day, but how likely will it actually get there in the next day? And could a service level that is generally tied to a five day SLA actually get there within just one day?

And with this information you can actually determine that again, by using our AI generated predictions, you are leveraging the billions of shipments that we've seen so that you can have that confidence that yes, this ground shipment will get there within one day. So in the past it's kind of correct me if I'm wrong here, it sounds like our data and our analytics were able to pull up that information and you could have dug through to find, you know, maybe which ones were the right options.

But this adds like a conversational aspect, just sort of makes it easier to access, is that right? Like what is the conversational piece with it? You start with the conversation right at the beginning. Maybe even before you start with the conversation, you might have an internal conversation to say, what are your goals?

What, again, I keep saying goals and I, I don't mean to sound like a broken record, but you need to figure out what are you trying to do with your logistics strategy first. And then once you have that clear objective, whether it's, hey, I wanna make sure my on time delivery percentage increases, or I just want to decrease my shipping costs, if those are your goals, great. Take those, and then you come to Luma and you help Luma get you to those goals. So you would have a conversation.

You would say, hey, I'm trying to decrease my shipping cost. What should I do? Luma can then recommend what are the next steps to get there and not only recommend what are the next steps, it would also tell you how much you can decrease your shipping spend by a specific number, so then now you actually have like a key result that you're trying to achieve. I love that you say that there's different goals.

I know you said you're harping on it, but I don't think so. I was just reading an interesting article recently that talked about how cost has been, you know, the be all, end all number one goal for companies for a really long time. But we are seeing a little bit more of people looking at, you know, your service level agreements. We're seeing more people looking for stability of carriers.

You know, as carriers have kind of come and go and, and looking for all kinds of different things when it comes to carriers, it's not just costs that matters. And so I love that you're bringing up, you figure out what your goals are and then come and bring that to the conversation. I think it's a great place either, actually, I think for me it's easier to see it. So I think this would be a really great spot.

I know you have a video that you've put together, kind of showing Luma AI Advisor. Is it okay if we go ahead and pull that up on the screen so everybody can see it? Yeah, let, let's jump on in. This video is actually a classic example of why someone would use Luma Advisor and our shipping optimization tool.

In this video, what you would see would be someone trying to see what the effects of adding a new carrier to their mix could actually bring. And so here someone says, hey, what happens if I add Amazon to my carrier mix? How does that affect my parcel spend? How does it affect my time in transit metrics?

And so in here you'll see, cool, select adding Amazon as that new carrier. And then what someone would've done before we implemented the LLMs is they would have to go through this entire dashboard and really look through it to figure out, hey, what actual benefits am I getting from this? How, how does this actually help me meet my goals? Now all they have to do is go and type into the LLM, hey, how much do I actually save by adding Amazon to my carrier mix?

And it will immediately pop out that answer specifically for you so you can then decide, oh yes, this is worth adding this new carrier. Ah, I gotcha. And so are those the kind of questions that you're seeing customers actually using with Luma ai? Those are probably the more advanced questions.

All, although once someone gets there, they're start asking those questions a lot. Users also ask basic questions. Like, what is the best carrier that has the the highest on time percentage? And it can instantly answer that for you, or what carrier do I spend the most on?

Or what weight do I end up shipping the most of so that I can try to optimize for that specific weight of a shipment. Could you ask like per zone? So could you say, hey, what, in zone three, what is my typical weight? Yes.

You can say for zone three weights between five pounds and seven pounds, how much could I actually improve my parcel spend by adding a specific carrier? Mm. Yeah. I love that.

It. Do you feel like it's more valuable for certain type of shipper, you know, high volume enterprise, SMBs? Anybody? You know, is, is there a key, key audience here?

I would say it is across the board. The use cases tend to change as volume ends up increasing. But let's say you're a brand new shipper. You're, you're doing, let's say a hundred orders a month.

What this could tell you is, cool, you're doing all your orders on just one carrier, most likely. This will let you know, hey, should I invest in building out another carrier integration? So one, it can help answer those questions so easily. Then as you get larger, this helps you really fine tune your shipping strategy.

So, hey, you know what, now I've grown, I do 10,000 shipments a month. I really wanna hit that two day SLA. How much do I have to invest into my shipping operations to meet those two day SLAs for my customers? What about like international shipping?

Is that something that it helps answer questions for? Luma Select, our rate shopping tool does help address international shipping. And so it will heavily rely on carrier estimates when diving into time in transit stats. But, for our dashboards, our dashboards are highly focused on US domestic.

Okay, great, great point there. So if you are US domestic, small, all the way up to large it can be a really great tool for you. I love that you mentioned Luma AI Select. I know that you know, cross border and, and international shipping is really complex and really challenging.

So any tools that help with that, that we do have Luma AI Select. So please make sure you're looking at that. What are kind of results, I guess, are we seeing early on? Can people expect cost savings?

Fewer exceptions? You know, what are we seeing so far? We're seeing a lot of cost savings, mostly due to users integrating with a brand new carrier that they weren't even considering before. And what Luma has actually highlighted to them is just as I've been saying, those savings that they could achieve if they added those different carriers and now that they have added them, they are already seeing those savings.

There was a recently an article that came out. And I'm sure for all of you who watched our trends up so recently, killed me, that said last year at this time companies were keeping about 13% of the increase in costs that they've been experiencing over the last year for themselves and giving most to customers. Well, as things have shifted over the last year, and customers are a little bit more wary, that's now at 40%. So costs are really hurting shippers, and so I can see the cost savings anywhere we can, especially these are easy savings that, you know, it can improve your service level.

You know, it can improve your customer experience as well. Completely makes sense that the cost savings is a big deal. I had to imagine, Tyler, that it would kind of adjust maybe the way shippers are thinking about their overall carrier strategy using tools like an LLM. Are you seeing any of that now?

What, what could people expect in terms of kind of their overall strategy shifting if they start using a tool like this? The biggest thing that I've been seeing is people are more open to having a multi-carrier logistics strategy. In the past, I would say not even that far back, like three years ago, the majority of our shippers, I mean the good majority of our shippers were just shipping on one carrier. Now that they're seeing all these benefits, they're opening the door to say, hey, you know what, I need more options.

And they are seeing those benefits come in, in the forms of savings and on time percentages. Yeah, I, so cool. And it really, really makes sense. I think everybody knows from a risk management standpoint, having only one carrier is bad.

From a cost perspective, having one carrier is bad. But it's always just been really complicated and really hard to do. And so I love the technology's getting us to the point where we can do these things and we can make huge differences without as much risk. Okay.

Tyler, where do you see Luma AI Advisor going in the future, what, can you give us a sneak peek into what we can expect over the next year? Yes, absolutely. So immediately, what we're working on right now is having the ability for Luma Advisor to not just be dependent on the data that's presented on the current dashboard that's someone's viewing, but have the ability to really just answer any analytical question that you ask it. That's coming out in the very near future.

After that, we do hope to expand Luma Advisor to also help answer questions like, how do I enable a new carrier? And also have more information around carrier metadata to answer questions like, which carrier is the best for shipping HAZMAT? And which of those carriers is the easiest to integrate with that does that? So having all that information in the backend of Luma Advisor is something that we're working on now, and I can just see how beneficial that will be once that ends up being rolled out.

It's gonna be huge. One thing I really love, we're right on the cusp of LLMs in the shipping industry. We're just at the beginning, and as you become part of it, especially if you are an EasyPost customer, please take advantage of this. You are there and kind of in the early stages of giving feedback and our product team is amazing at taking customer feedback and using it to implement things.

Are there questions that you wanna know? Is there data you wanna know? This is a great chance for you to get in and, and to be part of that. One thing also that I just wanna share that I think is really neat that sometimes our customers miss.

We have so much data at EasyPost on hundreds of carriers, so many service levels. We have billions of shipments of data and information. So you could go, you know, maybe there's an LLM or, or there's a chat bot you could use through a single carrier or through a smaller service provider. They simply won't have the same amount of data.

So if you go to a one of the big carriers, they only have data on their own services. They don't have data across carriers. If you go to a newer company, there's a lot of other companies that have LLMs that just don't have the dataset that we have. So critical to getting accurate, actionable information.

So Luma AI Advisor, it is so huge, so incredible. Tyler, if people want to learn about it, what is their next step? There's a couple things they can do. One is they can definitely visit the EasyPost.

com products page and check out Luma AI. That can give you just a very brief overview. If you are currently an EasyPost user today, you can navigate to the Luma self-serve to sign up for Luma yourself and actually start utilizing it as soon as today. And then lastly, you can also contact EasyPost sales rep and they can give you a nice demo of the tool itself as well.

Awesome. So many ways. Feel free to reach out to me, reach out to Tyler on LinkedIn, anything you want, you know, somebody's gonna help you get there. So thank you Tyler, for being here.

I am, this is one of my favorite product updates we've ever had, having an LLM specific to shipping. Huge. Thank you, thank you, not just for being here, but for making it. Absolutely.

Thank you. And I'm sure we will see Tyler again. Fan favorite. Everyone wants to hear what's going on with EasyPost.

All right. We'll see you all next time. All right. Bye y'all.