[00:00:00] Lori Boyer: Welcome, Unboxing Logistics family. It is so great to see you as always. I love our Unboxing Logistics community. You are all some of the smartest, brightest, hardworking people in the logistics and supply chain industry, and it is such a pleasure to be with you. As always, you know I’m your host of Unboxing Logistics, Lori Boyer of EasyPost.
Okay, family. We’re gonna be talking AI. You know, we talk AI every other second. You hear AI coming outta your ears, AI coming outta your nose. But there’s also a reason that it’s, and that’s ’cause it’s big, it’s huge, and it is disruptive and making massive changes in the industry.
Well, on the flip side, being used only in specific places where it’s working. So we’ve brought someone in. We’ve got Dana Moffatt here from Acumatica. We are gonna be talking about AI in operations. Dana, will you introduce yourself? Give our little audience here a little background of who you are and where you work.
[00:01:01] Dana Moffat: Sure. Thank you Lori. I’m a product manager for distribution integrations at Acumatica. And a little bit about my background. I’m a product builder at heart. I love the creative process of software development. I work closely with our team at Acumatica to gather business requirements, articulate what we need to build, why we need to build it, and who we’re building it for.
I’ve spent my career solutioning for complex distribution and healthcare supply chain, and I’ve been with Acumatica for almost five years now. I live in Montreal, Canada.
[00:01:32] Lori Boyer: I love that. I heard you say process.
[00:01:35] Dana Moffat: Yeah, I was that the giveaway?
[00:01:38] Lori Boyer: Love it. Love it. That is fantastic. So Dana, I would love to hear, you said you’ve been at Acumatica five years.
Also fun that you’re in healthcare. I was in healthcare in a past life as well, so that’s really cool. Two complex industries. Who is somebody that you kind of admire from this industry, or even just in your professional career? Can you share somebody who you admire?
[00:01:59] Dana Moffat: I think the person that I admire most is Ali Janney, our former Chief Product Officer at Acumatica.
He’s just retiring now after 15 years of being with Acumatica, and what I admire most about Ali is that he has this customer centric product design focus. So we are always solving real world problems for our customers. He encouraged us to get out in the field, visit our customers so that we could see and hear their stories.
Conduct user research and bring back those learnings to our development teams in Acumatica. So helping us to understand those customer opportunities and challenges is really the key to building great product that our customers love. And we discovered as we were doing this that our customers really loved to be part of that process.
They like providing us feedback. They like showing us what they do, how they do it. We can understand them better and build better capabilities in our ERP, which is an enter enterprise resource planning product. Ali also championed our Acumatica Community. It’s a virtual meeting place where our customers and partners collaborate on product questions and new ideas.
So that’s what I really valued about Ali is that product leadership and me and mentorship. And I just really like to thank Ali for all that he’s given to our product team over the years.
[00:03:17] Lori Boyer: Shout out to all the Alis out there. It is, every one of you pretty much watching or listening wherever you are likely have customers, and being customer centric and looking at customer problems, you should never be the hero to everyone else.
You are looking at the customer, figuring out ways that they can be the heroes in their own life. Your product should always revolve, whether it’s shipping, whether you’re a, you know, retailer, whether you’re ecommerce. Look to your customer, see what needs they have and work to address those needs.
Don’t just try to fit a round peg in a square hole. So I love that, Dana. That’s fantastic. And shout out again to Ali Okay, let’s dive in. I wanna talk AI. I am a super big AI nerd, Dana. I love AI. I love all the excitement that comes with it. Everyone here knows when they hear me, I love to talk AI. But let’s start.
Everyone basically is being like, okay, yeah. What are you doing for ai? What’s your AI strategy? What’s going on? From what you’re seeing, you know, what do you feel like is really going on in the industry? How much of this is actually progress? How much are people just like, everyone else has an AI strategy.
I need an AI strategy too. I’ll just make one up. You know what, where is kind of the flavor of the industry that you’re getting right now?
[00:04:37] Dana Moffat: Well, from what I’m seeing with many of our customers a lot of them have already been experimenting and exploring AI, and they’ve been doing it for a while now, either personally on their own or in their businesses.
I visited customers that have been experimenting with these AI technology technologies to solve specific challenges like automation. So they were trying to automate manual processes in order to maintain headcount and get more done in the business without having to have more people added on.
And then they would realign their people to focus on higher value work. So that is real progress. They were already looking for ways to leverage AI to solve real business problems. Then they came to us with ideas and challenges for us to incorporate it into Acumatica ERP. So I see that they really do have focus, but they weren’t looking to solve every problem with AI.
Only those that made sense for them at that time. So it wasn’t seen as an answer to every question or problem. Although we should ask ourselves before we do anything, can AI help me or short, shortcut this for me? And in Acumatica, our approach is really pragmatic. I would say it’s focused on our customer needs and not just AI for the sake of AI.
And I think that’s important. Although granted, if your competitors are exploring and already using AI, that’s a competitive advantage. So if it helps ’em automate, gain, operational efficiencies, you don’t wanna fall behind and you don’t wanna be late to adopt that technology. So in that sense, there is that pressure I would say, Lori, not to fall behind.
And I think that’s what’s great about Acumatica and our EasyPost teams, is that we collaborate closely with each other and our customers. And that includes our AI initiatives. So both of our products are already leveraging these AI technologies as part of our joint solution feature set. That’s what we do best.
So that frees up our customers to focus on the business of running their business.
[00:06:34] Lori Boyer: That’s fantastic. There was a couple of really key things I thought you said in there, Dana, that I wanted to repeat. I loved how you mentioned that it’s not to solve everything right now, right? Like, look for specific problems you have.
I hear that again and again. I’ve interviewed so many AI experts and that is a key element there. You need to have a problem that you’re gonna try to fix and not just try to grab AI and put it in anywhere that maybe you didn’t even know you had a problem. But I also loved how you talked about it being a differentiator and that there is that real risk of falling behind if people take off.
Where are you seeing it actually work right now? This is probably the question I get more than anything else. You know, we have a lot of disparate systems out there. You go in the warehouse, people may have 20 different technologies they’re using. They don’t connect well. The numbers don’t cross over well. And, and it is a little bit of a struggle right now.
So what, what are you seeing? What, what activities is it really working for?
[00:07:37] Dana Moffat: So where I think it’s really working well right now and where people are getting a lot of value is using AI as for the purposes of anomaly detection. An anomaly detection is like is like having another pair of eyes at the back of your head.
So you have AI looking for adverse trends or situations to call your attention to so you can act and course correct. And I give you a perfect example from Acumatica. Our solution has something called margin anomaly detection. So it helps you to detect if you’re in danger of not meeting your margin goals.
So in an era of rising costs where margins are already thin, and that’s holds true for a lot of our shippers. It can call out situations for you to act on. So I think that’s really helpful. I would say another great idea that we’re working on is collaborating with EasyPost. So we’re wanting to bring AI enabled rate shop capabilities to our customers in the future.
So I think that will have a very real impact in the future as we collaborate together on that capability.
[00:08:42] Lori Boyer: I love it. So anomaly detection, massive. Really, really big. We have so much, so much data in this industry, so having tool that doesn’t get tired and doesn’t let things glaze by, really, really great.
I love how you mentioned EasyPost, Acumatica and other companies out there. So if you are working, you know I mentioned earlier in a warehouse maybe have 20 technologies. A lot of companies are developing AI things within maybe solutions you’re already using out there. So it could be you need to be looking into the vendors and the partners you’re working with.
They could have some things available that you’re not even aware of. I know it’s hard to keep up with that, but rate shopping. I wanna back you up on that one. There is, to me that is one of the best use cases to be able to just quickly look through and figure out. That’s been kind of a, a bear for a lot of people for a long time. But figuring out which carrier is gonna be most reliable in which lanes and at which times and at which costs, and that is an easy no-brainer for me.
Dana, on the flip side, where do you feel like maybe we’re a little overhyped versus what our capabilities really are right now?
[00:09:54] Dana Moffat: I think there’s a lot of hype out there about AI as the answer to every situation or problem, and that it’s going to replace humans.
[00:10:03] Lori Boyer: AI’s not doing my laundry yet, Dana. It’s not.
[00:10:07] Dana Moffat: No, I, I hear you Lori, and you know, I wish I could do things like that. But I mean, in our day-to-day workplace I think there’s concern or worry about that and, and maybe there’s a lot of hype behind that.
You know, I have to admit, I was a little bit skeptical at first when we started to use AI tools as a product manager. I started to use an AI assistant. And, and I saw, while she’s not really replacing me, she’s helping me to summarize meetings, conduct research, accelerate my day-to-day. I don’t really see her as a threat, but she’s a helper, a tool that makes me more productive and efficient.
And I think that’s the same for, you know, our listeners today. If we focus on AI as a helper or an assistant that we can leverage to get more done and not see it as the, the answer or the solution for everything or, or as a threat I think it could be very helpful and it could really be a competitive advantage for a lot of our customers.
[00:11:08] Lori Boyer: Yeah, you’re pointing out something here that’s kind of quietly not always spoken. Sometimes, people are afraid to implement AI because we’re afraid to be replaced. And we hear about it in a general sense, like, oh, it may re, you know, replace workforce. But on a personal level, sometimes we feel like we need to prove that I am more valuable.
And it’s exactly like Dana said, once you jump in and start using AI, you realize this is not yet at the point where it’s really working without me. Now, maybe five years, we’ll see what happens, but right now it really needs that human guidance, but it is an incredible assistant who doesn’t care if you get mad at it and who you know is gonna be working 24/7 for you.
And really can help with a lot of that kind of busy work. I think that’s critical, Dana, don’t you think especially with labor challenges have been massive in our industry forever. And in a way this really helps bridge some of those gaps. Rather than replacing, it helps, you know, alleviate some of those pains.
Have you felt like that as well?
[00:12:14] Dana Moffat: Yes. I mean, I’ve heard that especially during COVID when, you know, a lot of workforce kind of contracted, if you will, and warehouses had to do, make, do with less employees. So there’s challenges in, in getting workforce into the warehouse, into the shipping desk.
And if you have tools that can really help the people that are there you can do more with less. I think that’s a win. But I think that’s an important point is AI is not here to replace people, but the people weren’t there. There is a contraction in the workforce. We have these very real struggles.
It could be a really important tool for our customers to leverage to make up for the shortfall in labor.
[00:13:04] Lori Boyer: Absolutely. And you know, that’s actually a global trend. I was recently reading this really interesting article about how China’s big focus on AI is not necessarily to try to get ahead, but to fill in the fact that they have such an aging workforce with their one child policy that had, and they really just needed to help assist with human labor.
In a way, I think that’s true for us across the board that there are tons of opportunities with AI. Don’t be scared of it. Don’t hold off learning how to use it really well, just because of that. Okay. So we do hear Dana of a lot of companies who may invest in AI and then it doesn’t go well, people don’t implement.
I’m curious to hear your thoughts on maybe why sometimes AI doesn’t deliver or you know, are there red flags that people should look for? I’m like, maybe I shouldn’t invest now or in this way, you know? What, what tips do you have around, I guess, trying to make sure when you invest in AI that you do get good ROI?
[00:14:01] Dana Moffat: Yeah. I think you, you have to have clear objectives and goals and that you’re gonna be measuring key results. Like any strategy, I think you need to go know, going in what you want to accomplish and why. And if you don’t have those clear objectives and, and you’re leveraging AI in situations where it doesn’t make sense, that could cause you to question your investment in AI.
So I think it’s really important that you, that you know your why. Why do we think AI is the answer for this particular problem or challenge? How will we know that AI ha has helped us transform or achieve our objectives? And how do we measure success or failure? Then benchmark that against pre AI.
So another challenge, what you kind of touched on is aligning your employees as well and motivating them to use these AI technologies. So there’s this element of change management that I think is, it’s always a critical success fact factor. As and, and that go, that’s hold true for ERP implementations and, and I’m sure for EasyPost implementations as well.
Having a product champion is always an important ingredient. So champions lead the way, they advocate for the new way of doing things. They’ll highlight the benefits of using AI and it’ll help your employees grow and focus on more meaningful and value added work, I think.
[00:15:22] Lori Boyer: Okay. I love that you mentioned change management here.
So we have some interesting trends around AI. A couple of things. I think it’s like 95% of employees are using AI even on their own, what we often call shadow AI, where people are using AI tools that weren’t official. Even though, you know, only 27% or something of logistics companies have some official AI programs in there.
One of the challenges with AI is maybe if you’ve got people who have been using tools on their own, getting them to switch over to the official tools that you are implementing. You know, for instance, I had on my own used a bunch of, you know, ChatGPT things and then our company kind of switched over to Claude and that was a little bit of a challenge for me.
So I think you are absolutely right. Get those champions out there. Get people. It’s not just necessarily adopting new AI. It can be switching from tools that they have been using even on the down low, that shadow AI over to tools that make sense because it’s got your entire background in there. You know, when we get shipping specific, you get Acumatica, you get EasyPost, you get other tools that are based on your shipping data.
That’s always gonna be better information. And so you want to convince people to kind of switch up their processes, move over, get alignment. If you don’t get alignment, if you don’t get input. There’s a good chance that it’s, that will flop.
[00:16:49] Dana Moffat: Yeah, totally agree. And, and like you say, you don’t wanna go it alone.
There may be a, a product out there like EasyPost or, or Acumatica, that have already solved the problem that you’re looking to solve. So you know, you don’t have to start from square one or we reinvent the wheel. You can leverage that work that they’ve already done because they’ve already invested a lot of time and resources and money in evolving these features for you that maybe you can just leverage out of the box.
[00:17:18] Lori Boyer: Or as you mentioned, Dana, that you said, you know, when you were talking about Ali and loving to get customers involved in sharing, reach out to some of the vendors you work with, reach out to the different companies and say, hey, this would be a really cool AI idea. Are you doing this? Maybe it’s on the roadmap. Maybe it’s something that you can help with.
[00:17:38] Dana Moffat: Exactly. I think communication is key and collaboration is key, especially with customers because you, you are the subject matter experts of your domain and you can bring us those use cases.
If we don’t have something that is an AI capability now that covers a use case that you’re wanting to be covered, bring that to, bring that to us. So we invite our customers to bring us these use cases to collaborate on our community and we will often have interviews with our customers or go visit them to get these requirements and bring them to life in our product because we wanna deliver capabilities that bring value to them, that solve real problems.
So it really is this, we need you as much as you need us.
[00:18:24] Lori Boyer: Yeah, completely, completely agree here. And on, on a similar note, look at your workforce as well. Like where Dana mentioned, you know, we’re trying to get that change management, go to them and say, what areas do you see where we could use a little more ai?
Where do you think that there are opportunities? And then maybe reach out. And it could be that there are those people that you’re already working with or, or maybe you need to get a new tool. So let’s, let’s talk Dana, about layering in maybe AI. Do you feel like, are there certain things that they need to have in place operationally before they start adding ai?
[00:18:58] Dana Moffat: So I think it’s really important to know your processes to have standard operating procedures or to have clearly defined frameworks. So that AI can be really leveraged to the best of its ability. So if your procedures or workflows, they’re constantly changing, or your processes vary by team or customer, no standardization, it’s hard for AI to learn from patterns because you don’t have patterns.
And if humans can’t explain the process clearly, well then AI can’t fix that. I think that’s a good call out. So I think that’s really something important to think about when you’re layering in AI. I think also thinking about your use cases. So you may have something that is cool, but it’s a low impact use case ’cause you haven’t tied it to your objectives and your key results.
So they’re gonna fail to show you that return on investment. So that would be, I think, chasing AI for the sake of having AI. So I really look to see is it solving these real world problems that bring value to the business?
[00:20:07] Lori Boyer: Okay. I love that you said that because honestly, some of the best use cases for AI are not very sexy.
You know, it’s like we wanna go for the glitter that this sometimes backroom accounting, so really boring stuff is beautiful use case for AI. So don’t get caught up all in the glitter and glam of some of these. You had an amazing demo, but is it actually gonna bring you ROI Look at what your actual use cases are.
That was spot on Dana. I love that. Okay. If somebody’s gonna be getting started, do you have recommendations? What would be the first steps you would take to try to, you know, add in AI or if you already have some AI, and you’re wanting to expand, you know, is that different? Do you follow the same steps?
What? What are your recommendations?
[00:20:56] Dana Moffat: I think it’s really important to look at your data and make sure that you’ve got good data, because I want you to remember that, that when you’re training AI on bad data or that data hasn’t been corrected it’s gonna use that to inform your decision. So this could result in you incorporating bad data into these models and perhaps lead you to make less than ideal decisions and may cause you to think that AI is failing me.
So remember, AI can inherit bias from historical data and it can reinforce bad patterns if you don’t monitor it. So I think monitoring it is really important and, and, and, and remembering, well, what are we training the AI on? So really, I would say try to, as much as possible, clean up your data before you start leveraging that AI.
It’s a matter of, you know, garbage in, garbage out. So let’s make sure that we take out the garbage and, and, and, and have clean data before we start. So I think that’s really important to look at, at that and also to look at, well, what, what source of things are we training this model on? So what to look at, what not to look at.
So remember I said, you know, humans are still needed. So you’re right. I heard you say that. We are the guide, right? So we tell it what to look at and what we’re interested in. So that guiding exercise is still important. So that’s important before you start to really kind of think it through and make sure that you’re starting with the best data possible.
[00:22:25] Lori Boyer: Yeah, absolutely. I’ve heard it from a million times, your AI is only gonna be as good as your data, both the data it’s trained on, like Dana said, and the data that you share with it. So even making sure that you’re working from a single source of truth, you know that you’re not having different numbers in different areas that different teams are using.
All of that is a, a big thing. Love that. That is exactly perfect. What if you’re adding more AI, do you have any tips around that? So let’s say that you’ve added AI maybe in an area, but you’re thinking you could do more. Do you think of expanding from that one area? Do you add in different systems?
I, I know there’s not necessarily a perfect answer for this, but I’m just curious what you think.
[00:23:11] Dana Moffat: Yeah. So again, I think you have to look at well, what’s my why? So if I’m expanding the workflow, is that workflow well understood? Is it something that has patterns? It could be well articulated that AI can learn off of?
So I think that’s really important. I think also that when you tackle AI, you don’t want to treat it as one lump sum project. Tackling projects as smaller pieces and measuring the results and the outcomes and, and recognizing that it’s okay to make mistakes and to have failures because those are opportunities to learn.
You’re learning. This is all new to us, right? And that experimentation and exploration, I think it’s very important. So as you expand into the business the different areas or opportunities for AI, I think it’s important to pause and kind of assess, well, what, what did we get from AI in this particular part of the application or the business challenge that we’re solving for before branching out and expanding it in the business?
I think that’s a worthwhile thing to do is to pause and reflect and then reassess your goals before you branch out and, and determine, well, where is the best bang for our buck, if you will. Where are we getting the most return on investment for this AI capabilities that we’re investing in?
[00:24:35] Lori Boyer: Really good advice here is to be moderate. You know, it’s like, it sounds like you’re saying let’s move forward, but not so fast that we’re just throwing everything in all willy-nilly and don’t know what’s going on. I also really liked how you said that again, to not be afraid to try things, because even if it messes up, we just learn from it.
I always say that mistakes are an opportunity to learn, or I tell my kids like, thank you so much for giving me so many chances to practice patience. You’re so good at letting me be patient, but it’s true. You know, all of these roadblocks can actually be something that turns into something really great.
Do you feel, you know, we’re kind of talking on the, a little bit of the challenges side here. Are there things that you think companies kind of believe or, or fundamentally think about AI that maybe is wrong or even that makes their operations worse? Have you seen AI initiatives, I guess, that have just really flopped Any lessons we can get there when we’re talking about learning from our mistakes?
[00:25:34] Dana Moffat: Yeah, I think the thinking that AI reduces a need for humans or that it can replace humans, so I think that would be a mistake. I think there’s still a very strong need for domain expertise and especially in shipping and, and in, in ERP, enterprise resource planning. So we have domain experts in these different functional areas, and they’re very important, so AI cannot replace them.
These domain experts help to define problems for AI. They interpret its outputs and they validate its results. ‘Cause remember, AI can hallucinate. We still need the humans. So it’s not like we’re going to put AI on autopilot, autopilot, I would say. We still need these domain expertise. So if we’re thinking that we can just throw AI at any problem and that we can replace human beings, I think that is a mistake. And that’s what AI and operations where it can just go wrong.
AI can help us sift through all that data, help us to interpret the data and ultimately arrive at better decisions quicker. But I mean, humans are still in the driver’s seat. I mean, AI capabilities will evolve. They’ll get stronger. But we’re the ones that are still driving what we use the AI for and interpreting it as results and identifying critical areas where we can leverage this capability.
So I think that’s really important, Lori.
[00:27:00] Lori Boyer: I 100% agree on that. I had actually recently talking to someone whose company made this very mistake. Really, they got rid of a whole bunch of a department thinking they could switch it with AI, and it was a disaster, and now they’re trying to rehire people, bring in temporary workers.
We’re just not at that spot yet. As you said, it may evolve at some point, but right now there are no AI technologies that I have seen that can fully replace humans in our industry. So be really cautious about that talent and the humans behind it is such a valuable resource.
So if we’re gonna be cautious somewhere that is really good advice. I’d love to hear we’re just about out of time, but I would love to hear from you two things before we go. First, tell us about what Acumatica is doing with AI and maybe just a little bit of overview so people can be excited. And number two, any final advice? If people were gonna walk away today and do one thing, what would you want them to do to embrace AI or, or change the way that they’re looking at things?
[00:28:04] Dana Moffat: So you’ll see from release to release Acumatica is bringing a lot of AI capabilities to our customers. So we’ll have automation, so I mentioned the anomaly detection, the pair of eyes in the back of your head looking out for you to catch certain situations so you can course correct.
So we already have that in our product. AI assistance. So providing you assistance while you’re working, for example. So you can ask it questions, it will look through your Acumatica data and provide you answers. I think we’re also looking at ways that we can bring our customers into this development cycle.
So we do do a lot of outreach to our customers. We visit them we do user research and we involve them in these AI initiatives. So a lot of our AI initiatives that we’ve developed, it’s been with the help of our customers who have been these early adopters, providing us with feedback so that we know that we’re solving the right problem for our customers and bringing value to them.
So. We follow our own advice. We don’t just throw AI at anything. We use it very focused way to solve pragmatic problems for our customers. So we really are concerned about bringing value to them and solving these real world problems.
[00:29:27] Lori Boyer: Yeah, I I love that. EasyPost is a really similar mindset. That’s why I think we’re great partners. But AI is becoming integrated in such a way that it’s just part of EasyPost. You know, it’s not like this add-on, buy on, oh, and here’s a side AI product. And I think that’s really how life’s gonna be moving forward. You’re in your, your platforms, you’re in your work, and there’s AI helping fuel that and run things in the background so that you can get the answers. I see that as where it’s going in the future.
So, really cool opportunities. Dana, so, one piece of advice.
[00:30:03] Dana Moffat: So I’d say start with one to two high impact, low risk use cases. Make sure that you set those OKRs, which are the objectives and your key results. Look for something that’s repetitive, measurable, and already structured. ‘Cause remember we had talked about that, so we wanna avoid those mistakes.
So a good example would be like shipping delays, your ETA prediction, optimize carrier service selection, anomalies, and using AI to assist you in decisions. So that’s where I would start. I would clean your data. For that first use case so that that model isn’t working with that garbage in, garbage out that we were talking about earlier.
And I’d also measure your outcomes and iterate. ‘Cause remember, you’re gonna make mistakes. You’re learning, right? And if you’re not making mistakes, you’re not learning like you had said Lori. So I think that’s really important. Celebrate your successes and treat failures as a learning opportunity, not a reason to abandon AI.
[00:31:02] Lori Boyer: Yes, I, that is so, so good. Everything is perfect there. Start small. But boy, when those wins start adding up and you’re celebrating, it does give you kind of a high and you’re excited. Like my eyes have been open to all the crazy new possibilities that there are with AI. And that is the fun, fun part of AI.
So Dana, thank you so much for being here. This has been fantastic. What if somebody wants to learn more about Acumatica or if they wanna connect with you, if they have questions about AI, how could they do that?
[00:31:34] Dana Moffat: Sure they could go to acumatica.com and learn more about us there. You can also join our Acumatica community.
You can be a guest. Sign up as a guest. And you can also reach out to me at dana.moffat@acumatica.com.
[00:31:50] Lori Boyer: Perfect. That is fantastic. And community, keep working on it. Don’t ignore AI just because it might be a little scary sometimes. Dive in. It is exciting. Thanks again, Dana.
[00:32:03] Dana Moffat: Thank you, Lori.
[00:32:04] Lori Boyer: Bye-bye.
