The Power of AI in the Warehouse With Henrik Bergsager and Kushal Pillay From Powerhouse AI - Ep. 46
In This Episode
What comes to mind when you hear the term AI? For some, artificial intelligence is an exciting step forward, bringing possibilities of error reduction and cost savings. For others, it’s little more than a marketing buzzword.
Henrik Bergsager and Kushal Pillay of Powerhouse AI are in the first camp. In this episode, they discuss how AI tools can significantly improve warehouse operations for small and medium-sized businesses.
Uses of AI in the warehouse
AI can be used to improve virtually any warehouse task, including inventory management, demand forecasting, and storage optimization. Henrik notes that businesses should approach problems one at a time, not all at once. “By focusing on these smaller cogs in the warehouse, you don't need massive amounts of investments to get up and running.”
Getting started with AI
Getting started with a new tool is always daunting. For a smooth transition, Henrik says, “You have to get the rest of your employees on board. … You need to get all the warehouse workers understanding why we're doing this, [what are] the benefits of introducing new technology.”
Kushal adds that organizations should focus on the KPIs they’re trying to improve with the technology. He adds that it’s important to keep expectations realistic—don’t expect the tool to eliminate human labor. “[At the] end of the day, you're not buying a robot. You're buying a certain software, which does a job for you.”
How much technical expertise do you need?
Fortunately, you don’t have to be a technology genius to take advantage of AI warehouse solutions.
As Kushal puts it, “Most of the tools out there [don’t] need you to know the deep details of how they work. … The only thing we need is what is the input, and what do you expect as the output? The rest of the work is done with[in] the system.”
Links
- Connect with Henrik on LinkedIn: https://www.linkedin.com/in/henrikbergsager/
- Connect with Kushal on LinkedIn: https://www.linkedin.com/in/kushalpillay/
- Visit Powerhouse AI’s website: https://www.pwh.ai/
Transcript
Lori Boyer 00:00
Welcome everyone to Unboxing Logistics. You know the drill. I'm Lori Boyer, your host, here from EasyPost. And today, we are going to be on the subject of, as you all know, one of my favorite topics, which is AI. Love AI, and we're going to be talking specifically about AI in the warehouse. So this is gonna be a really, really fun discussion.
Everyone buckle up for this one. Get your notes out. You're gonna want to be paying attention fully for this one. But, to get started, I've invited a couple of amazing guests, who know so much more than I do about this topic, from Powerhouse AI. So I'm going to throw it over to them to introduce themselves.
Henrik, let's start with you.
Henrik Bergsager 00:50
Yes. Thanks. So thanks for having us, Lori. My name is Henrik Bergsager and I'm the head of partnership and sales here at Powerhouse AI. And a little bit about me, I would say like for the last 10 years or decade or so, I've been working at startups that develop solutions for regulated industries.
So for example, like distributional pharmaceuticals or perishable goods.
Lori Boyer 01:13
Really cool. Kushal.
Kushal Pillay 01:14
All right. Hi, I'm Kushal. I'm the director of Powerhouse AI. We have been running this company for the last three years now, and we have got into the space mainly because my background is in robotics and my vision was always to get software or robotics in the hands of humans.
And that's how we started this company. And that's how we are here today.
Lori Boyer 01:38
It is so fun, Kushal. I love, first, I love startups. I also love robotics. I go to a trade show and I'm instantly trying to find all the robots and all the cool stuff that's going on. So I am really excited to have you here today. Okay.
I'd like to get to know you just a little bit. So I'm going to ask you a couple of questions, and then as we dive into our topic, I'm going to ask each of you to give us one or two kind of takeaways. So it's really important to me that our audience if they have to jump off or whatever the thing is, that if they only take away one or two things that they can remember from today's discussion, what it is you want them to remember.
So I've been asking all our guests this season what their favorite comfort food is. So when you're having a long day, when it's just been whatever it is, because, you know, are you going for some pizza? Are you looking for some sort of warm drink? What, what is your comfort food you go to?
Henrik Bergsager 02:42
For me, I would say like so I'm in New York, so I would say greasy Chinese food. I think that's, that's a good comfort food.
Lori Boyer 02:51
Greasy American Chinese food, right?
Henrik Bergsager 02:54
Yeah. Yes.
Lori Boyer 02:54
Awesome. Okay. I love that myself. So good. Kushal?
Kushal Pillay 03:00
For me, it's Mexican. I usually go for tacos. That's my go to food usually.
Lori Boyer 03:05
Okay. I love some good tacos, some chips and dip, all of that. Okay. Kushal, can you share with us a takeaway you have on the topic of AI in the warehouse. What, what do you want our audience to remember from today?
Kushal Pillay 03:22
I think one of my biggest takeaways, which I want everyone to take away, even our, even the people I meet in warehousing and everyone is, especially with the AI sense, is to start small with clear objectives and scale gradually.
That's one of the key takeaways I usually emphasize a lot on throughout my journey. So that's been one of the key insights. The second thing would be not to complicate things too much, not to make it too complicated when you're starting off it, and because that leads to my third takeaway, which is like not having clear objectives.
And you end up not having clear objectives on what exactly you want to get out of this.
Lori Boyer 04:05
Okay, those are all fantastic points. And to me, honestly, if that's all you hear. Especially we're talking kind of smaller entry level, kind of small to med size, middle, medium sized warehouses, so critical to become overwhelmed really, really fast.
Henrik, what do you have to add?
Henrik Bergsager 04:25
Yeah, I think I want to like play on what Kushal said. And I think compared to like in the past, implementing like AI doesn't necessarily mean that you have to transform like your whole warehouse completely. I think AI tools are becoming more like focused, kind of like what Kushal said, like, like focusing on specific operations or, or resolving a specific issue.
So they're all becoming a little bit more plug and play. So it's easier to kind of integrate with like your existing systems and it's, it's more affordable and it's faster to roll out and see value. So I guess my takeaway is that kind of don't be scared of kind of approaching AI because it's, it's definitely like the threshold of adoption is definitely lower than what it used to be.
Lori Boyer 05:13
Okay. I, I love that. I think that when some of us think AI, we just immediately think dollars. And hard and really complicated. And so that's super interesting. So I kind of referred to the fact that today's discussion is going to sort of revolve around that small to medium size kind of warehouse operations.
Henrik, would you mind explaining to us? Let's define that. What does that look like to you?
Henrik Bergsager 05:40
So when we talk about like, when we talk about small and medium sized warehouses, we typically refer to like operations, I would say between like 10,000-50,000 square feet. And they would be handling, like, I would say moderate volumes of inventory.
And in terms of their function, their core kind of like, their service, I would say like more like regional supply chains, where you have maybe like some manual processes as part of your warehouse and some semi automated processes.
Lori Boyer 06:13
Okay, Kushal, do you have anything to add to that? Is it usually just a single warehouse they would have or could they have multiple?
Kushal Pillay 06:19
They could have multiple. Usually it turns out to be having local goods, not like international goods. It's like local delivery goods, which is more confined within a single state or a single city. That's how it, that's how usually a small local warehouse works.
Lori Boyer 06:35
I've been in multiple local multiple warehouses. I love to go on warehouse tours whenever I can. I don't know why, but they're so fascinating to me.
Henrik Bergsager 06:44
I like that too. I like that, too. For some reason, it's just fascinating. Yeah.
Lori Boyer 06:49
I miss this. So the smaller ones, I feel like, are often a little less tech friendly. You know, a little more elbow grease sometimes have been put into it, and, and maybe the processes aren't exactly the perfect process in place.
So, when I see these warehouses, I think a lot of times they don't automatically think. You know, maybe I'm off on this. Maybe it's not how it is, but do you feel like that perception is changing? You mentioned that AI can be a good fit here in starting out. What, what is the reality when it comes to AI?
And I'm going to start with Kushal on this and then we'll hit Henrik. When it comes to this kind of size of warehouse.
Kushal Pillay 07:36
So when it comes to AI and with respect to size, I think it comes to my first learning like usually when it starts out with small warehouses, people believe that it's, it's oh, it's a technology that is not fit for us, you know, we have to start, you know, from the scratch.
It's because it's also the industry on, on the ages of which has been running for so long. That it doesn't, they don't really land on, like, okay, AI can, AI is not someone, first of all, they have a misconception. AI it's something that it's, it's not there yet. It can reach there, but it's not there yet. It's an, AI is an assistant.
That's how it should be seen. So whichever task is mundane and it's repetitive, such tasks should be focused upon and that's where it should start off with.
Lori Boyer 08:21
Okay, I love that statement. AI is an assistant. That's how it should be seen right now. It is not yet at the place where AI is here to replace your labor.
So warehouse owners, runners, operators think that first, especially in this small to medium size, you're using AI as a tool to improve then Kushal, is what you're saying, kind of improve the processes and make the job easier. Henrik, is that, is that what you would agree with? The purposes?
Henrik Bergsager 08:52
And just like your operation, just more cost effective because, because AI tools are more like agile now, so you can kind of target like the operation that is important for you to automate. So I don't think we should think about kind of AI as, okay, I have to like implement like full scale robotics. I think you can focus AI on tools related to like, let's say, inventory management, or like, let's say that demand forecasting, or if you just want to use it for like optimizing your storage. And by kind of focusing on these like smaller part or smaller cogs in the, in the warehouse, you don't need these like massive amounts of investments to kind of get up and running.
I think that's, that's, that's important. I think that's also contributing to that, the perception is shifting. Yeah.
Lori Boyer 09:41
So start small is what we're saying. Like I have seen and again, my nerd mind gets blown, but you know, the fully automated warehouses where you're just like, well, what is happening? Am I living like a hundred years from now?
But for our audience today, that is not what we're talking about. So start small. I love how you threw out a few of those kinds of areas. I'd love to dive into the really specifics. Where are the AI opportunities? What are the processes or tools that should be used at this stage? Kushal, do you want to start with maybe one or two, give us examples of where you can use AI.
Kushal Pillay 10:18
Yeah. So like something to start small of it, I would. I mean, what Hendrik just said, like, demand forecasting is definitely one thing that I would really recommend to start off with. That's something easy and it's easy access and it's always been around, like, Power BI, it's such a simple tool, like, everyone uses it, most of them use it. Even I guess, automating any repetitive task, even, like, invoices, deals, delivery orders, like, there are. Even ChatGPT can do it.
You can just upload a delivery order on ChatGPT and it can transcribe it into text for you where you can just upload it to your WMS system. So it's, it's an easy way to quickly see quick wins, small returns and small changes. Like that's, that's the best way to start off with, which is scalable also for your own processes.
Lori Boyer 11:06
Okay, I love that. Demand forecasting, automating things like your invoices, your delivery orders, using things like ChatGPT. I always any in any discussion I have on AI. Sometimes we just need to get familiar with using tools like ChatGPT. So if you're not using it and you need to just kind of get used to it, start using it and in regularly and you'll get more comfortable.
Henrik, what else? What else can we use?
Henrik Bergsager 11:31
I think, I think at this point there's so many options out there and there's a lot of like the startup scene when it comes to AI tools are just there, it's flourishing right now, so it's, there's a lot of options. So I think, I think you can kind of look at it. Okay.
I have this specific issue that are very much like pertaining to how I operate. Let's say you're handling third party labels, right? And you need something to be able to read those labels. But you don't have a tool. You don't want to like redo your whole barcoding platform. Then you can kind of find a tool that does that and helps you kind of like plug that gap and provide you with that data.
So it's a very, it's more customizable now. And I think that's really interesting. It's really exciting.
Lori Boyer 12:15
That's really, really cool. So number one, what I'm hearing here, identify an issue you have. So maybe it was you mentioned around third party label scanning.
Henrik Bergsager 12:26
For example, yeah.
Lori Boyer 12:27
So, so walk me through that example. So let's, you know, I'm not exactly the warehouse expert here. So what kind of challenge that might you be having with third party label scanning? And then how would you then, you know, what kind of AI tools do you feel like there are to help with that?
Kushal Pillay 12:42
Yeah, so for example, with third party labels, right the most simple ones is mom pop stores, which we call it in Southeast Asia, where these small vendors are sending certain goods, selling their goods through big e-platform providers, right?
They're trying to sell goods out there. So they have their own labeling system. They have their own labels put on the product, which is not well configured for that warehouse. So a certain warehouse where it's going for, like, for example, let's say you're selling it to on Amazon. So you're going to send it to the warehouse, sending it to the warehouse that they're going to distribute it from that they have to relabel it whenever it comes in, which so that the assistants can start reading it. Right. Imagine this for a small medium warehouse. That's not possible. That's like too much hard labor. They have to reprint everything. So now with AI, we can actually use the same label to create a virtual label, which like stays behind it and stores information for every warehouse that it moves towards.
So you don't have to change the label physically and it works with the same label that the mom pop store has put it in. So that's, that's like the basic example of how things can work.
Lori Boyer 13:54
So Henrik, that brings up a great point. You mentioned there's a lot of startup kind of companies out there who are offering a lot of AI options and products like this.
So, so let's say that we did have a company, we're going to keep on this example here, that exactly like Kushal explained, you know, they were having this label challenge. So what was the first step for them to try to figure out where they might find an AI product or solution or, you know, and this could be for anything that they were having.
Maybe they're having demand forecasting and that's going to be an easier one, but, but maybe they've got a niche problem like this. Recommendations for people to go out and research where to find these kinds of things and what's available.
Henrik Bergsager 14:37
Yeah. When it comes to like SaaS platforms. Especially like when in like distribution, like a TMS, WMS barcoding system.
There are portals online that can help you find the right software specifically for you. You can, like it's simply, you can go to like websites like Capterra, SoftwareAdvice, G2. These are all platforms that kind of gathers all the softwares out there, and they'll provide you recommendations based on your need.
And they're very much like customer friendly, like user friendly, so you'll be able to like talk to a person, explain what your challenges are. And they can kind of come up with some recommendations on, okay, this platform might solve that problem for you. So I think that would be a very, that's an easy way to kind of do some research.
Lori Boyer 15:29
That's a great recommendation. I was going to say, I often just start straight with the Google search, but you know what's kind of funny. Kushal, I was thinking this. You could even use ChatGPT to probably put in, hey, do you know of any solutions that offer AI around these kind of tools? And then absolutely exactly what Henrik was saying.
Kushal Pillay 15:50
Every warehouse runs the weed in a specific manner, so the use case of the process of how it works, they're looking for different solutions, but even if the process is the same, like, for example, in a warehouse, it's one process which is inbound where they receive the goods and they import it within their warehouse.
That, that same process is hugely different, even if the industry is same. So, it's very custom, and that's why the platform, which Hendrik suggested, works, because, you know, it's, you have a specific solution for your use case to how, so that it works perfectly for your finding.
Lori Boyer 16:26
Yeah, I think that's a great idea, and, you know, a lot of you work with partners all the time.
This is something I hear about in this industry so much. You may have partners who would have ideas of people they know of. You know, if you come to EasyPost, for instance, we work with so many different vendors and groups and, and If you just ask your CSM, if you, you know, have you heard of any tools that work?
You can get a lot of great ideas that way too. So Henrik, what are the objections? I think a lot of people feel scared then. So let's say that somebody now finds a great tool. It looks great. But I don't know. There's like a fear sometimes on implementing new things. What, what are the common barriers that you see when people are trying to implement AI? How can businesses kind of overcome those?
Henrik Bergsager 17:18
I think like the first and foremost. That kind of like I see that are maybe applies for a lot of the smaller ones that hasn't really adopted technology like this in the past is that one thing you kind of have your like your management, like they want to solve an issue.
They want to gather data on certain processes, right? I think first and foremost, when you're kind of incorporating and adapting technology into that something that is new. You kind of have to get like the rest of your employees on board. So I think like education around like why you're introducing these new technologies. Let's say, and you're in a warehouse environment, like you need to get all the warehouse workers kind of understanding why we're doing this, like what's the benefits of of introducing new technology.
Lori Boyer 18:03
That's great. Anything you have to add there, Kushal? So, getting everybody on board especially with the idea of we're not necessarily, we're not here to replace you, we're not getting this technology to replace you, but to help you. Anything else that helps them feel comfortable onboarding new technology like this?
Kushal Pillay 18:21
I think when they start focusing small, the key things that we start giving emphases on when they're adopting a new technology would be the KPIs. I think that's, that's like a very important factor when you're understanding on which technology you're trying to adopt. So what is the KPI you're trying to improve for your business with this technology?
For example, with the label scanning solution that we were speaking about earlier, right? The third party label scanning solution out there, the key, the key answer for the KPIs would be accuracy and efficiency, right? It should be quicker and less error prone, right? That's the key issue that they are facing.
So with the adoption of certain technology for that use case, they should measure, okay, what is their expectation with this? And what, it should not be unrealistic expectations that, you know, it should work, it should do everything for me. And I should not spend, I should spend zero human hour in this. No, because end of the day, you're still not buying a robot.
You're buying a certain software, which does a job for you. So it should have realistic expectations with what you want to achieve. And that will also help them, you know, see quicker results, right? Because that's what I want to see. You don't want to invest in a technology which you're going to say it's going to come off in two years that you're going to have in return.
You're going to have it in a few months. That's, that's the key intention I would want them to focus on.
Lori Boyer 19:39
Okay, a couple of huge truth bombs dropped there by Kushal that I need to make sure that we reiterate on. So just because something is shiny and beautiful and exciting. If we can't like track its KPIs. And if we can't be able to see what kind of ROI we're going to get on this tool.
It probably means that that's not a great investment for you to make right now. The second thing that I thought was huge that you said was the fact that we shouldn't overestimate. You know, this is going to be the silver bullet that changes everything in my warehouse. And so, Henrik, Kushal, both of you, what do you feel like are realistic ROI that people should be anticipating?
Of course, this is going to vary massively from, you know, one warehouse to the next, whatever the tool is. But are there any guidelines you can give our listeners out there to kind of understand? Is a, is a 5 percent improvement something that would be good? When you're implementing AI, should you expect that within a week, everything's going to be changed?
You know, what kind of guidelines could they look at to even know that they're right on, on track?
Henrik Bergsager 20:57
Yes, I think it's, it's, it depends on what your starting point is. So where we see like, often you have like a drastical improvement, like a really good like return of investments is like the elimination of human error related to something.
So let's say you go through your process, your operation, you see like, okay, we're struggling of, captioning correctly, like the count of boxes on a pallet, for example, and that is impacting our ability to have like, okay, what's our inventory? So by introducing like AI, you can have like 99. 9 percent accuracy on the count.
So you're eliminating that human error aspect. So the, the ROI is like very much like it's fast. You can see it right away after implementing it.
Lori Boyer 21:44
So human error. So it kind of goes back. Okay. I think that, I don't remember which one of you said earlier, the fact that you need to look at where your personal problems are going on.
If you're having human error issues, you can see some massive improvement there. So human error kind of across the board, do you think most areas of human error in the warehouse, there are some sort of tools to help with that, or, or do you think there's anything that they should say, well, we're just in trouble in that area.
Henrik Bergsager 22:10
I would say there's tools that like, like the elimination of human error or not elimination, but the minimization of human error. That's one of the main advantages of, of AI. Like the ability to gain control of your operation and gather all the data without exceptions. I think that's, that's one of the main kind of advantages when you start early that you can kind of see quick returns on.
Lori Boyer 22:37
Kushal, anything around ROI that you can share?
Kushal Pillay 22:40
So based on my experience, I guess what I've seen, the key areas where most warehouses should start off with would be order accuracy, which is something that we spoke about, human error. Second would be inventory turnover. So like every inventory how long it's staying in your warehouse. And how long it does. It makes a huge difference because you are your warehouse is a space utilization You're renting space to a certain good. So it depends on which kind of warehouse you have, and then it's also with cost reduction.
So that's everything, right? Like that's a number of hours you spend, number of reworks needs to be done. These three are the key main elements that I've seen where usually an investment or the ROI factor makes a huge difference with AI because how AI works, it's it's basically keeping your experience with you, no matter what task it's doing.
So that's what it does. So what I mean by that is, if you put it in layman's language, it would be, if you give it to an experienced operator, or if you compare it to an intern, if you give the technology to any one of them, it's going to perform exactly the same. It's going to give you the same accuracy.
And it'll improve. The more you use it, the better it becomes because it works on feedback. And that's where it makes a huge difference on your operations as well.
Lori Boyer 24:00
I love that. One, it made me think as you were talking about the different team members, one of the questions I get a lot from people, especially when we're getting a little bit on the smaller side, is about what kind of tech staff is necessary.
So if, if a warehouse is thinking of implementing new tools, you know, all of the excitement around data analytics. AI, all of this recently, you know, are we talking people need data scientists on their team to analyze data? On the smaller end, I guess, how tech savvy should companies be in order to start implementing basic AI, or if they're working, say, we'll say even just with the Powerhouse AI.
You know, what kind of tech support available is that I think that scares a lot of people.
Kushal Pillay 24:48
Yeah. In all honesty, I think with respect to the customer, I would, or the warehouse itself, I would be, you just maybe as savvy as you're excited about using ChatGPT. Like that's the basics we need. And I think that's enough for you to get into the space of AI because most of the tools out there doesn't need you to know the deep details of how it works. It only needs you to, the only thing we need is what is the input and what do you expect at the output? The rest of the work is done with the system. So that's what is the beauty of it. Because previously that would be where consultancy used to come in, right?
Where we have to sit with you, understand with you, build a whole system for that specific warehouse and then starts functioning. So that part is more taken out with this.
Lori Boyer 25:40
To me, what I'm hearing is exactly why I think I personally love AI. AI's purpose is to make technology easy for the stupid people like me who don't know all the background and, and really brings all that data and that knowledge.
Makes it accessible for just the everyday person like me. Is that kind of what you were saying there, Kushal?
Kushal Pillay 26:07
Yep, that's exactly what I was saying.
Lori Boyer 26:08
Okay. So that's perfect. I might not be as smart as Henrik and and Kushal, but I can jump in with AI and you should be able to, so I think that's a, a flag as well to look forward when you are looking at different tools.
Is this easily understandable? Is this something that you shouldn't need to be hiring data scientists for in order to get it? Any tips. You know, we've talked about look at a problem in your warehouse, but are there any kind of, Henrik, low hanging fruit or easy things to kind of get your feet wet when it comes to AI in, in, in this kind of small to mid sized warehouse realm?
Henrik Bergsager 26:46
Let's say you're just starting from scratch. I would choose whatever system you decide to do. Let's say you want to get like a WMS system, right? Or a TMS system. Choose a cloud based one. So that the data that you're collecting is more easily shared so that when you do decide to kind of incorporate some third party AI tool data is what's all about, like that's what matters.
So then if you have that in a place that's easily shareable you can then faster get value out of whatever AI tool you choose.
Lori Boyer 27:22
That's great. What about if you're thinking of growing? So one of the other things that I've seen, I was speaking with somebody recently at a warehouse that gotten a tool that they loved, but then they kind of outgrew it.
You know, as they scaled and started growing, they ran into the challenge of they hadn't kind of thought ahead. And thought, oh, what, what will I need in five years? Any suggestions, I guess, for people when it comes to looking into the future and knowing I guess making sure their, their AI grows with them.
Kushal Pillay 27:54
I think first thing, like you've already taken the best step forward. Like when you start off with it, if you're building, if you're using a software platform, which is scaling with you, with its own API is having better solutions for you. And if you see that trend, and if you don't see growth in it, that's, that's a red flag where you should automatically start understanding because if they're not growing with, even with your current state, that means they are not growing beyond your current state also. So when you are there, they won't be ready. That's, that's the whole point. Like, so they need to be two steps ahead of you. And I think that's the flag you need to keep in mind all the time. Like, so for example, if you buy a software service, which is handling as simple as an Excel sheet, right?
Like, which is there. And then you're like, okay, now I'm in the stage where I can't handle everything on an Excel sheet. You need to move that away for sure. So because an Excel sheet doesn't change anything, it's just columns and rows and formulas on it, right? But how can I further maximize my usage with respect to users, access, and so on and so forth, if that doesn't exist?
So you'll have to use this software, which is always two steps ahead of you, so that you are ready. The software is ready when you are ready.
Lori Boyer 29:05
And I think as, as I think it was Henrik mentioned earlier, there are a ton of startups out there. And the fun thing about working with startups is that they are so receptive to customer feedback and hearing what kind of tools are needed.
So in this industry, I cannot emphasize enough the importance of creating good relationships with all the people you're working with, with your, you know, whether it's your WMS, whether you've got a shipping solution, whether, whatever it is that you're using, create those relationships and be vocal. I am the worst sometimes at sharing when I just like to complain about problems and I say, ah, my, my software doesn't do this instead of just asking.
So I think that that is exactly right. But, you know, look for somebody who's going to work with you, grow with you. Henrik, what are you most excited about as we move into the future kind of of AI? What do you feel like are the next like big technological advances coming?
Henrik Bergsager 30:08
Yeah. So before I jump into that, I do want to kind of like add to what you just mentioned.
I think that's so important, Lori, is that for anyone who's looking to kind of adopt AI, like work with startups. Like they would be so much, they're, they want to work with companies. They want to get feedback on their solution. They want to learn how their, your operation is working so that they can learn more about the space so they can develop their own platform further.
And often what you can do is that if you work closely with a startup, like you can get, I guess, insight into, like, what's coming in the future in terms of the pipeline. What are they developing? Where is the platform going? So that way you will have a better take on whether or not that kind of, like, correlates where you, where you're going, like, where your operation is going.
Lori Boyer 30:55
I, I wanted to say as well, when you are kind of a smaller in the field, so you may have a smaller warehouse, working with a startup is so smart because when you go to the huge massive corporations, you're not getting discounts. You're not, you're, you're too, you're small fish, right? Where when you work with them, they're so excited and, and you really do feel like a big part of it.
Henrik Bergsager 31:16
Yeah. If you're open to kind of like taking that, like everything is not going to be as structured as a working with a big corporation, but you also will have the opportunity to have a say. In terms of like, you'll get some, you'll, you'll be able to give input on what's working for you. And maybe that will have like, that will maybe translate into their, their product being a little bit more customized towards how you're working.
So you can get some advantages of working with startups. Definitely. Yeah, no, you had my original questions about like, what, how I see the future and like AI and supply chain and what, what there's a lot of out there. So it's very like. There's a, there's a lot of kind of like to, to talk about, but what I think is interesting now is AI and, and, and how things operate is often dictate a little bit about guidelines and regulations, like state guidelines or like local guidelines. So for an example, what's interesting to see now you have like the FDA has come up with new guidelines that's coming into place in 2026 about the handling of perishable goods, right? Like how is this going to be stored, distributed and so forth. And that will kind of push everyone in the direction of adapting technologies that can give them that kind of control and security and provide that data and being compliant.
So I think those kind of advancements are, are, they're very interesting. And I think that's going to be a big driver on where the technology is going and how the industry is, is adopting new technology to be compliant.
Lori Boyer 32:58
That's such a great point, because I have had people say, so we often talk about the sustainability aspect and environmental stuff that's going on in this industry.
It's so big. And, and a lot of that also gets driven by regulations and things that are coming forward. But some things that I know companies are doing is they look to countries that are a little bit ahead of like the US. So for instance, Kushal, you're in Australia, right?
Kushal Pillay 33:22
Yes.
Lori Boyer 33:23
Often we may look to Australia or we may look to Europe or you know, some of those regulations come in a little bit earlier.
Do you feel like you see the same with AI or technology or any of those things? Are there places that people can look to to kind of keep an eye on, well, this is happening there now. So maybe in five years it might be happening here and I should start getting things in places.
Kushal Pillay 33:46
I think it's hard for AI at this stage because I think to be honest that it's the first regulations that are coming on AI is related to the finance industry.
That's where it's starting because that's the most tricky industry to implement something, a technology that learns by itself and continuously grows by itself so that it has to be controlled in a fashion where it's still regulated within human rights, right? So, but that's Europe is the first place it's been implemented and they are implementing such rules.
With respect to the general use of AI and how it's been adopted. That's mainly based, currently, it's based on state law or government laws of that country as this stage of today. But beyond that, I wouldn't. It's a concern, yes, but it should be a concern to a stage where you are doing something beyond your control and at this stage, it's not there yet, at least with the current technology.
So it's an assistant, as I said, so it's not going to take over anything as of today.
Lori Boyer 34:49
I think it's still a really great piece of advice. Let's maybe look outside of our industry. There could be other industries that are seeing advances more quickly or, you know, different regulations even, more quickly than we are.
And so look at those and kind of keep an eye out on on what might be coming. Okay, we're almost out of time. So I want though for you to share to our audience a little bit about who Powerhouse AI is and what it is that you offer as well. So that if they're, you know, if you're listening today and you are wanting to explore some AI options, these, this is an absolutely amazing company.
So I would love to hear from you. Whichever of you wants to take over this first, go for it.
Kushal Pillay 35:31
Yeah so we at Powerhouse AI, what we do, we are basically building the next generation scanner. When I say that, that means it's the Google Lens for warehouses. So you take a picture of anything in your warehouse, it'll know exactly where it is supposed to be.
Is it in the right quantity in the right place? That's what it does. So with just a single picture, you can understand everything about your goods and it keep track of everything that you get in your warehouse. That's what we do.
Lori Boyer 36:00
That's, that's super cool. Henrik, what are some of the most common questions you get about it?
Henrik Bergsager 36:03
I think it's questions about when we talk to potential customers is like, how does it incorporate with our existing systems? Like how flexible is it? What are the limitations? Those kind of things.
Lori Boyer 36:16
Okay, great. And I think your first question, that's when I hear constantly, integrations with other technology and software.
So where, if people did have questions, they wanted to learn more about you, learn more about your scanner, different things that works with AI, where should they go?
Kushal Pillay 36:34
So they can just hit www.powerhouseai.com and they'll have all the information they need.
Lori Boyer 36:40
Okay. Perfect. If people are interested in connecting with you personally, you know, are you, are you two on LinkedIn?
Are there areas where people could kind of follow you, learn from you, kind of get those sort of insider insights into what's happening with AI? Henrik, we'll start with you. Are you on a how can people reach out to you?
Henrik Bergsager 37:02
Yeah, I'm on, on LinkedIn and feel free to kind of reach out just to kind of talk about AI and warehouses and. Yeah, it's just it's an exciting space to be part of right now. So yeah, just sharing information and getting some use cases. That's always interesting.
Lori Boyer 37:17
Perfect. Kushal, what about you?
Kushal Pillay 37:18
Absolutely. Even I'm on LinkedIn. You can just search my name, Kushal, Kushal Blake, be able to search me. And I would be happy to talk to you or even understand if you have any situation you are in, in a confused way, what to use exactly to start off with.
I would be happy to talk to you and help you out of there.
Lori Boyer 37:35
Absolutely. Again, as we mentioned earlier, these gentlemen were so good about reminding us to reach out to your partners, ask people if you're trying to find solutions, a lot of people know what is going on. So this has been a really helpful discussion, really good specific takeaways.
I appreciate that so much for those in the warehouse industry. The biggest takeaway I had was just start. Try something, start small, don't get too crazy and think you're going to create an entirely robotic warehouse to start. But find something and become part of these opportunities. So anything else that you want to add?
I want each of you have a chance to say goodbye.
Henrik Bergsager 38:19
I, I think just, just, if you kind of just are very new and you just want to learn about AI, I would, I would look into like, there's a lot of like AI courses on platforms like Coursera and edX that you can, kind of get inside and learn a lot from. And as we mentioned, like LinkedIn, there's a ton of LinkedIn groups there within every and, and like every niche little market so join those groups.
They usually have webinars that are posted there. So you kind of listen in and learn and, and kind of see what other companies are doing in your space.
Lori Boyer 38:55
I love it. Kushal.
Kushal Pillay 38:57
I think the same advice goes here. Start small, start simple. You just go on YouTube, search something small, whatever you have an understanding of AI, start with that keyword and let, let the internet take you where it's supposed to take you and understand that on what AI runs.
Lori Boyer 39:14
I love it. All right. Thanks so much. And we'll see everybody next time. Bye bye.